PARENTAL ADEQUACY AND THE pORT

Parental Value and Adequacy as Reflected in the Perception-of-Relationships Test:

Empirical Data on Seventeen Statistically Significant Signs

Barry Bricklin, Ph.D.

Adjunct Associate Professor

The Institute for Graduate Clinical Psychology, Widener University

with special assistance from

Michael H. Halbert

Consultant to Mangement

Bala Cynwyd, PA

Abstract

This article summarizes approximately 40 years of research with the Perception-of-Relationships Test, representing a sample population of 1,508 cases. New normative, reliability, validity and interrater agreement data are summarized.

It also features our pioneer efforts to use a group normative reference standard, to add to the single-participant data that already exist. A single-participant reference helps one to understand the specific value a particular parent has to a specific child. The group reference used in this research allows one to red-flag when a parent may have negative value to any child.

Four groups of children were formed in which the parents represented different levels of caretaking adequacy. Groups I (n=16) included children from intact families. Group II (n=34) included children from parents involved in mild custody disputes. In both Groups I and II the parents showed no clinical or test evidence of poor parenting skills. Group III (n=40) featured continually fighting parents who used their children as message-carrying pawns. Group IV (n=40) represented seriously neglectful or abusive parents.

Twenty-three signs on the Perception-of-Relationships Test were hypothesized to appear differentially in these groups. Seventeen reached a statistical significance level equal to or less than .05. Several statistical and theoretical caveats were raised about using drawing-based research to formulate testable clinical assertions.

Drawings As Projective Tests

One goal of this paper is to show that drawings can be helpful in assessment despite their generally poor press (Garb, Wood, Lilienfeld, & Nezworski, 2002, p. 459; Thomas & Jolley, 1998). A second goal, an exemplar of the first, is to show that drawings can generate empirically-derived hypotheses about the adequacy of parents both from a group perspective, as well as from contexts of dyadic and other family systems. This approach allows an evaluator to assess not only how a child assigns value to caretaker A as opposed to B, which requires a single-participant reference, but also to recognize when a caretaker may not meet adequate standards of parenting from a normative perspective, which requires a group reference. Existing and new data collected for 40 years on the Perception-of-Relationships Test (PORT) will address both goals (Bricklin, 1989, 1992, 1995, 1997, 1998, 1999; Bricklin & Elliot, 1995, 1997, 2000, 2002a, 2002b, in press-a, in press-b; Bricklin & Halbert, in press-a, in press-b). Four caveats will be detailed. Drawings can yield clinically specific assertions only if precise access-code stimuli (defined later) are presented to the respondents. Second, there is a limited range of specific behaviors drawings seem able to predict. Third, it is difficult (but not impossible) to find validating evidence of these behaviors. Fourth, predictor models must be able to specify the system-specific contexts in which interpersonal predictions will be manifested. System complexities have created confusion in attempts to validate drawing-derived data. Few clinically specific assertions will be accurate other than within a limited number of contexts.

This paper is divided into two separate sections. The first section summarizes the existing research on the PORT. The second section, beginning with the sub-heading, “The Current Study,” reports the first use of the PORT with a group reference.

The Perception-of-Relationships Test

The PORT generates information relevant to family assessment. A core measure reflects the degree to which child-caretaker interactions lead to comfortable and effective behaviors on the part of the child. It consists of seven tasks, each on its own page. Task I asks the child to draw each parent. Task II presents a line drawing of a female figure and then a male figure. The instructions are “Here is Mom/Dad. Put yourself somewhere in the picture.” Task III has two line drawings of parent representations to the left and right of the page and the child is asked to put him- or herself in the picture. Task IV is a request to draw the family. Task V is a request for the child to draw his or her family “doing something.” Task VI has a horse between two renderings of homes, one of which says “Mom” on it and the other “Dad.” The instructions are “Here’s a little horsey. It’s time for the horsey to go into his/her house. Please draw a line into the house where the horsey will go.” Task VII consists of three cards. The first shows a little doggie asleep in a bed having a dream of its mommy. The child is asked what that dream is. The second part has the dog dreaming of the daddy with the same instructions. The third part asks: “ Which dream is nicer?”

Prior research shows that each task elicits associations of the child operating within a specific family system. Tasks I through IV reflect a child’s: overall relationship with each parent; being actually or psychologically alone with each parent; being in the real or psychological presence of both parents; wished-for ideal; view of each parent’s supportiveness of family unity. Tasks VI and VII reflect a child’s conscious belief systems, and allow a comparison between how a child responds when aware and not aware of what is being revealed. The scores yielded by the PORT will be described later. Validity data, have been gathered on 1,508 cases (Bricklin & Elliot, 2002a, 2002b, in press-a, in press-b; Bricklin & Halbert, in press-b). New test re-test data on 127 cases is also described.

Background of the Current Research

Between 1961 and 2002, thousands of PORTs were reviewed by our research teams. Some cases were litigated, most not. We had collateral and follow-up life-history information with each PORT, often collected over a 10-year range. Consistent relationships between certain signs on a PORT and other independent evidence of parental competence were noted. From information gathered between 1961 and 2002, it was possible to formulate hypotheses about the presence of these signs and independent evidence of parenting adequacy. Four groups were formed, one of parents not engaged in custody disputes showing no evidence of parental pathology or interparental hostility and the second of children from custody cases in which there were minor disputes but none involved parenting adequacy. The third group consisted of children of parents who exhibited intense interparental conflict (Bricklin & Elliot, 2000, p. 501). High-conflict caretakers, as they engage in continual battles, become poor parents (citations given later). The fourth group consisted of children whose parents had physically or sexually abused or seriously neglected them.

The PORT was originally designed to be used with a single-participant reference standard. A reference standard is the entity to which a measurement score is compared in order to derive relevance for a specific decision. It may be a previously examined group, a previously examined individual (the single-participant paradigm), or criterion-referenced (arbitrary expert opinion). (Measurement theory issues, the adequacy of the measurement units used by the PORT are discussed in Bricklin & Halbert, in press-a.) Since the PORT’s original goal was to compare how a child assigns value to Parent A compared to B, the single-participant reference was employed: a subject’s scores are compared to other of his or her own scores. The measurement goal was to understand a particular caretaker’s value to a unique and specific child. The current research uses a group reference. The use of both allows an evaluator to know how a child assigns differential value to each caretaker, and to see how each parent’s skills, as reflected in the PORT, resemble those of known groups of other parents—parents whose skills range from adequate to non-adequate. Before presenting the new data, existing data are presented to provide necessary background information.

PORT Data, 1961 to 2002

The PORT Predictor and Criterion Concepts and Measures

Independent validating criteria focused on the degree to which child-caretaker interactions lead to comfortable and behaviorally competent behavior in the child. How these concepts were operationalized is spelled out in Bricklin & Halbert, in press-a, based, in part, on theory and research best described by Bargh (1997), Horney (1945) and Zajonc (1980, 1998). A brief description follows.

PORT estimations of the parent with whom a child more comfortably and efficiently shared information were compared to estimations derived from extensive observations, often over several years. These observations were made by mental health professionals from criteria other than PORT scores. The original observation protocol is summarized. Emphasis was on (mainly nonverbal) ways a child demonstrated comfort and effectiveness during or immediately subsequent to interactions with a caretaker in spontaneous, structured and instructional contexts. The basic dimensions were a child’s movements: toward (positive); against (aggressive); away from (fearful). Research interest was on what impact parental behavior had on a child, not on what parents knew or did. Here are a few of the positive categories scored when a child was either speaking (initiating or responding), just listening, or listening and acting: relaxed and/or smiling facial muscles; leans toward other person; maintains reasonable eye contact; willing to ask questions; noting from whom the child most frequently and spontaneously seeks help. Following is the statistical profile, essential for an understanding of the validity data. Note that these data are attained with our team’s specific procedures. We begin with a short getting-to-know-each-other phase (about five minutes), followed by a 15-minute phase in which the child is presented with games and tasks slightly below his or her ability levels, to see how each parent behaves when the child does not really need help. This is followed by a one-hour phase in which games and tasks are used that are slightly beyond the child’s ability levels. This encourages parent-child interactions and hence magnifies our ability to observe who a child turns to for help, and how each caretaker responds. This phase is more highly standardized than the others, and is the phase that is scored.

In two samples, (n=60; n=37) the ages, ethnicity and SESs were directly similar to the norms given elsewhere in this paper. Each child was observed with both parents present, so he or she could make choices about initiated or responsive interactions. In a one-hour session, the distribution of positive scores ranged from zero to 12. At the outset of our research we tracked positive and negative reactions. Our research goals were adequately achieved if we counted only positive reactions, a practical “plus,” since fewer raters are needed. The mean number of positive reactions was 7.4; the Standard Deviation was 1.2. Six to nine positive reactions characterized 70 percent of the cases. A point difference greater than two represents a significant difference between two caretakers. The mean score of 7.4 positive responses per hour may appear low, but note that each child spent much of the hour immersed in task problem-solving, not interacting with the parents. These results are not comparable to protocols that count the number of positive and negative interactions initiated by parents (Lahey, Conger, Atkeson & Treiber, 1984; Kerig & Lindahl, 2002). Interrater agreement (three raters) was high, 90 percent, partly because the range of categories used was narrow: A>B; B>A; A@B; neither A nor B. The actual number of positive interactions noted by each rater achieved an agreement rate of 82 percent.

Legal decision makers render decisions in two main categories: legal custody (the right and responsibility to make key decisions for a child) and a time-share plan (or physical custody). In the former category, three choices are typically used: sole custody; shared custody; and very rarely, neither parent should have custody. (Nowadays, judges seem to favor shared legal custody, unless there are blatant reasons to exclude a parent from this role.) As for time-share plans, few judges venture beyond two choices: sole custody or joint custody. If the former is chosen, the non-custodial parent is usually given a boilerplate arrangement in which he or she has the child every other weekend, plus an overnight in the middle of those weeks when he or she does not have the child for the weekend. When complex time-share plans are awarded, they are almost always worked out by the parents and/or their attorneys, not by the data or the judge, and are usually based on parental-convenience, not best-interest, factors. This choice range has advantages in a research setting; it decreases hair-spitting arguments about how much better one parent is than the other. Thus, mental health professionals who offered validity designations used categories that reflect the narrow range of choices used by our legal system. In real life, mental health professionals can and do offer more complex plans. Newly emerging research (Kelly & Lamb, 2000) is encouraging the employment of complex time-share plans, but is considered controversial at the moment (Solomon & Biringen, 2001). Judges usually ignore complex plans unless the two sides agree on their own to implement them. Note that PORT data are more directly relevant to time-share plans than legal custody. Note also that it has never been claimed PORT data can adequately reflect an overall best-interests determination.

The formal statistical data follow. Data collected from 1961 to 2002 yielded a sample of 1,708 cases, 858 females and 850 males. The mean age was 7.76 and the SD 0.17 in a sample of 1,581 cases. These numbers were 7.87 and 2.101 in 127 cases. The SES was low-to-high-middle and the ethnicity 94 percent caucasian and six percent all other. The Parent-of-Choice or POC categories yielded by the PORT, matched to the range of choices required by the legal system, do not yield a normal or symmetrical distribution, so percent-of-agreement is a reasonable way to express our results, and are inherently more user-friendly to those trying to understand statistics in legal settings. Test selected POCs are never used to address ultimate issues or to assign a caretaker to a legal category. Overall, we have validity data on 1,381 cases evaluated between 1961 and 1997, and 127 cases investigated between 1997 and 2002.

The percent-of-agreement rates between PORT suggested POCs and those of independent experts based mostly on observations, tests (other than the PORT) and collateral interview data, are listed following the sample size, (the overall n=1,381). Structured task problem-solving by children with access to both parents, observed from behind a one-way screen by three psychologists (1961), n=30, 90 percent; courtroom judges (1964-1981), based on all data available, n=45, 89 percent; agreement with the Bricklin Perceptual Scales (BPS) choices (another data-based test, Bricklin, 1984) (1964-1981), n=23, 83 percent; courtroom judges (1981-1985), based on all data available, n=42, 95 percent; agreement with BPS choices (1981-1983), n=30, 84 percent; two psychologists, based on family therapy notes plus consultation with relevant therapists with families seen over two- to five-year intervals (1980-1985), n=30, 93 percent; courtroom judges (1986-1990), based on all data available, n=76, 93 percent; psychologists based on all clinical (except for PORT and BPS scores) and life-history data available (1995-1997), n=1,038, 89 percent. (Since judges may have been influenced by PORT data, the agreement rates between judges and PORT-derived POCs are offered as “information” and not validity evidence.)

In a current study (1997 to 2002), n=127 (Bricklin & Elliot, 2002b; Bricklin & Halbert, in press-b), we investigated the association between future and concurrent validity. Future validity indicated an 89 percent agreement rate between PORT choices for POCs calculated at Time-Point 1 with criterion designations made by independent mental health professionals at Time-Point 2, eight months later. Concurrent validity was assessed by comparing the PORT POCs obtained at Time-Point 2 with the criterion designations made at Time-Point 2. The agreement rate was 91 percent. The overall average percent of agreement between PORT choices for parent-of-choice and those from independent sources is about 90 percent.

The most recent test-retest data on the PORT over an 8-month period is shown in Table 1. TDS refers to Task-Difference Scores. The Task-Difference score or TDS indicates the difference in scores earned by each parent. A TDS of seven would mean that one parent had seven more points than the other. The frequency of each TDS in our sample population is given in the next two columns: the first indicates the frequency of a TDS at Test-Point 1 and the second at Test-Point 2. The other columns show how often there was a change in POC from Test 1 to Test 2. PORT results are quite stable over the time interval studied. Seven percent of the overall cases showed a change in POC, and about 66 percent of these occurred where the TDS was zero or one. If the TDS is greater than one, the likelihood that the test and retest scores will show the same POC is 97 percent. The likelihood of change of POC increases sharply as the TDS moves toward zero or one. A TDS of 0 or 1 shows a 21 percent chance the POC could change on retesting.

(Insert Table 1 about here)

Interrater reliability of PORT scoring was obtained from two samples of seminar attendees (n=36; n=41), in which more than half of the scorers had no prior experience with the PORT. Four different percent-of-agreement scores were obtained: (1) the points scored on Task I (the most complex); (2) the POC on Task I; (3) the overall TDS score for all seven tasks; (4) the overall POC based on seven tasks. The percent-of-agreement rates, respectively, were: 74; 90; 82; 92. Note carefully that the scale used to score an individual PORT task (or a single BPS item) is ordinal. The only information needed is rank-order, A>B, B>A, A=B. This scale choice was deliberate, because we did not want to overly influence an evaluator’s estimation of a given caretaker’s value to a child based on but a single life-area. Confusion is common here because the critical score in suggesting a POC, the TDS in the case of the PORT, represents an interval scale, and may be used as such. The use of an interval scale with either the PORT (or BPS) allows an evaluator to compare the PORT TDS (or the BPS Item-Difference Score) to how children in general differentially assign value to each parent.

The Current Study

Selection of the Experimental Groups and Clinical Hypotheses

Since this was our pioneer attempt to use a group rather than single-participant reference (to see if the PORT could red-flag various categories of parental adequacy), the decision was made to use experimental groups that were either markedly different from one another in representing parental adequacy, or, at least, in the degree of existing family unity. In other words, we were not, in these initial efforts, aiming to detect subtle differences. As is the usual assumption in our field, the thinking was that if our test predictors could not differentiate blatant differences among groups, there would probably be little hope they could pick up subtle differences.

Hence we conceptualized the criteria for inclusion in the groups in a way that it would be literally impossible for anyone who made an assignment of a case to a given group to be at odds with any other person making such assignments.

Briefly, one group was to consist of cases where the parents were married (i.e., the child tested with the PORT was part of an intact family), and there was no clinical evidence of poor parenting on the part of either caretaker.

The second group was composed of children from divorced parents, but where there was no evidence of individually poor parenting skills.

The third group was composed of children from high-conflict parents, who had been enmeshed in bitter disputes for at least two years.

The fourth group featured children from parents who had had their parental rights terminated, or who had been accused of substantiated abuse or neglect.

The formal criteria are given a bit later.

The hypotheses were formulated prior to group selection. The groups were selected to be matched in age, intelligence and/or age-correct educational placement, and socioeconomic status. Any child diagnosed with ADHD was excluded from this study, although prior research suggests this is not necessary since ADHD-children do not manifest PORT test-retest instability nor caretaker-selection bias (Bricklin & Elliot, 2002a, 2002b; Bricklin & Halbert, in press-b). Each group was formed so the children’s ages, intellectual/educational levels and the incomes of the parents resembled those in the overall population from which we have repeatedly drawn samples between 1962 and 2002. The ranges, means and standard deviations are presented following the descriptions of the groups. The assignments of families to the different groups were carried out by the author and at least one other member of our research team in conjunction with input from the mental health professionals involved with a given family, based on all of the clinical and life history data available.

The cases to be included were selected from among the thousands obtained between 1961 and 2002. The following conditions had to be true for a case to become part of the overall pool of cases that were subsequently divided into the four experimental groups. (1) At least one PORT was in the file. (2) Information existed on the child’s age, intellectual/educational and clinical status. (3) Information was present on parental income. (4) Clinical and life-history data existed that included complete personal histories of each caretaker and massive amounts of information on the history of each marriage. The latter included detailed investigations of each parent’s dyadic and family systems interactions with each child. The forms used to gather these data were published as parts of ACCESS: A Comprehensive Custody Evaluation Standard System (Bricklin & Elliot, 1995). Specifically, these forms are called: Child’s Access to Adult Parental Strengths (CAPS) (a detailed history of the marriage and child-parent interactions), Personal Life-History Data, Family Observations Interactions, and a variety of specific forms for pediatricians, educators, mental health professionals, significant others and collateral informants.

Following the formation of the overall pool, at least two members of our research team reviewed each file. All data were studied, except those from each PORT. As mentioned, it was our intention to set up highly discrete groups. Since the entrance criteria are sharply discrete, there were no disagreements between the researchers about which of the four groups a given file was to be assigned to. (The groups and other technical criteria of inclusion are described immediately below.)

At this point, the pool of cases was divided into the four groups described below. The next task, following the design of symmetric distributions of ages, estimated intelligence and SES, the groups were selected to be matched among themselves in these variables, and the means were engineered to resemble those in the populations we have repeatedly drawn samples from between 1961 and 2003. These norms are reported a bit later, and overall norms are reported in Bricklin & Halbert, in press-b.

Group I (n=16) were children examined for minor psychological reasons, mostly educational underachievement. There were two cases of posttraumatic stress disorder (a dog in one case and an auto accident in another) but none involved the quality of parenting.

Group II (n=34) were children whose parents were involved in mild custody disputes, centering on arguments over who could provide a better school, neighborhood, or extended family. All of the caretakers in Groups I and II met the criteria of the “good enough” parent. They encourage attachment appropriately and show an adequate understanding of children’s needs. More importantly, they display no serious negatives, for example, abuse, neglect, lengthy lapses in attentiveness or blatant episodes of losses of self control (Schutz, Dixon, Lindenberger, & Ruther, 1989, pp. 16-24).

Group III (n=40) consisted of children whose parents were involved in continual conflict, often within the courtroom, for two or more years. The involved children were “caught in the middle,” used as message carriers to deliver scathing messages from one parent to the other or to involved mental health professionals (Bricklin & Elliot, 2000, p. 501; Doolittle & Deutsch, 1999, p. 425-440). Research shows parental adequacy sinks to a serious low point when parents are so engaged (Bricklin & Elliot, 1995, pp. 38-40; Bricklin & Elliot, 2000, pp. 501-505; Hoppe, 1993, 1997; Hoppe & Kenney, 1994, 1995, 1997).

Group IV (n=40) consists of children in which the caretakers had either been threatened by the court with a possible termination of parental rights or actually had them terminated in the past and/or were accused of substantiated abuse or neglect. In order to match the parents in this group in socioeconomic status to those in the others, and the children in intelligence, the majority of these cases were derived from various private practices. Agency-originated cases tend to represent lower socioeconomic status levels. However, none of our data suggest that one gets different numbers from agency cases. Note that the PORT cannot be used in reference to any caretaker with whom the child has scant contact. Overall data suggest that caretaker-child contact is needed for several hours at least once every two to three weeks.

Since all of the PORT signs being investigated are negative in psychological implication, they were predicted to occur with greater frequency among the cases in Groups III and IV than among those in Groups I and II.

Normative Information on the Experimental Groups

The range of ages for each group was 6 years to 13 years. The mean ages and standard deviations are as follows. Group I: 8.9 years; SD = 1.87; Group II: 8.8 years; SD = 1.82; Group III: 8.5 years; SD = 1.56; Group IV: 8.6 years; SD = 1.66. The chi squared value (df = 21) is 2.032. The p value is .9999. The groups are comparable in age.

The range of incomes for the parents in each group was $30,000 to $70,000. The means and standard deviations are as follows. Group I: 57.4; SD = 9.60; Group II: 56.9; SD = 10.0; Group III: 55.2; SD = 10.2; Group IV: 55.5; SD = 9.50. The chi squared value (df = 21) is 7.032. The p value is .9980. The groups are comparable in income.

IQ scores were unavailable, most evaluators having used abbreviated versions of standard intelligence tests. The rule of inclusion was that each child had to be in an age-appropriate grade, maintaining at least a “C” average.

The PORT Signs Being Investigated: Clinical Hypotheses

This research stemmed from the observation that certain PORT signs appeared with high frequency among the children whose caretakers experienced serious problems with their parenting skills. But since the data were gathered via comprehensive custody evaluations, we had much collateral information available, pulled from psychological tests, observations, interviews and documents. I was therefore able to link each sign to existing clinical data and theory. The clinical hypothesis are offered to suggest to clinicians what each sign might mean psychologically, and to suggest foci for future research. While some may seem speculative, there are empirical data and theoretical information available to support them. Much derives from the field investigating pre-wired motor functions in human behavior (Calvin & Bickerton, 2001; Deacon, 1997), the research on how body-experience shapes almost all psychological functions including reasoning and the attribution of meaning (Cozolino, 2002, p. 147), and the evolutionary-genetic psychological forces that may prompt human drawings (Dissanayake, 2000). A major purpose in presenting this research is to show that while young children do not have the words to describe some of the complex perceptions I will ascribe to them, they do have inborn structures that allow them to experience and express such experiences motorically.

Some of the signs were predicted to occur only in relation to a particular caretaker (signs 1, 2, 3, 4, 5, 7, 8, 9, 10, 12, 13, 14, 15, 18, 19). Signs 6, 11, 16, 17, 20, 21, 22, 23 were considered “correctly predicted manifestations” regardless of the caretaker involved, that is, the problem could exist in multiple family systems. Some signs may appear in either context. This differentiation helps detect when a particular caretaker is probably contributing more to the problems than the other. The terminology in italics are the names used to identify the PORT variables in Table 2.

1. The child draws Two heads on a single figure, one usually grimacing and the other relaxed. It has been found when the access-code figure is a serious substance abuser, but at least adequate as a caretaker when not under the influence. The clinical hypothesis is that the child experiences the parent as a “dual” personality. Note that these signs may reflect psychological conditions that are perceived by the child not only as “bad,” but also as confusing or betrayals. A child may be more “hurt” by a parent who had functioned at a 90 percent level in a child’s perceptions, who then sinks to a 50 percent level, than by a 50 percent parent who becomes a 40 percent parent. While original PORT data, using a single-participant reference, focused on how a specific child assigned psychological value to one parent compared to another, with a group reference one can investigate whether parental behaviors, as reflected in the PORT, seem hurtful to children in general.

2. The child draws Two figures when only one is needed. For example, the child, directed to “Draw your Mom,” instead draws one figure then crosses it out and draws another.

3. I first ran into this sign at a workshop by Emanual Hammer (on a date I cannot recall, at least 20 years ago). The respondent makes a slash or X mark on a figure drawing such that it penetrates both (lateral) body boundaries. Hammer reported that in adult populations this sign was strongly suggestive of suicidal tendencies. I thought to myself that it would be a “hard sell” in a courtroom to convince anyone that something that looked like a slip of the pencil could have such serious implications. This very situation occurred. A five year old girl had drawn a slash-mark through her figure on the PORT task that asks her to draw herself in proximity with her mother. (This mother later had her parental rights terminated in regard to another child.) I did not know what to make of this sign in a child, because Hammer said it was dangerous when drawn by an adult. I reported it as “possibly very important information” in court. The opposing attorney scoffed and said that such markings were “obviously accidental.” I asked him if he would respond the same way had a radiologist said that a tiny dot on an X-ray represented cancer. “Would you say to this radiologist: I don’t think it means that at all. I think it’s just some little dot that doesn’t mean anything.” (I have a hard time explaining to attorneys that the implications of drawings are derived from follow-up databases, not from clinical guesses.) Nevertheless, the judge also thought the hypothesis of danger was ludicrous and took no action. Over the next couple of weeks this child managed to walk in front of a car (and luckily sustain only minor injuries), fall into a neighbor’s swimming pool and almost drown, and a few weeks later, when everyone realized how right my warning was, pull down a boiling pot of hot chocolate from the stove over her head, badly scalding her. (She had to wear a hairpiece for many years.) From a more empirical perspective, to this day I have only seen this sign where the respondent had some strong need to “X” him- or herself out of some situation, sometimes life. Interestingly, our data suggests that such signs, for example, the X-mark, derive from innate “deep structures,” (Noam Chomsky’s term, defined later) and not from learned sources. By “learned source” we refer to a situation where the child has seen comic book use of X-marks to indicate negation. We have observed such signs in the PORTs of children as young as 2 ½.

4. A severe Vertical misalignment is scored, for example, when the head is placed in a vertical dimension far removed from where the torso and legs are placed, even though the child finds some way to connect the head to the rest of the drawing. Severe misalignments seem to represent situations where a child has to bend him- or herself literally out of shape to accommodate interactions with the access-code figure. The child cannot relax and be whoever he or she really is; the child tries to find some way to accommodate interaction with the caretaker even though this creates physical and mental anguish. Clinically, we see it only in cases where the child has an ongoing and well documented conflicted relationship with the access-code parent. We have never seen it in a case where a child has a comfortable relationship with the access-code caretaker.

5. This sign is scored when the child places either his or her back to a parent or the parent’s back is placed toward the child’s front view (Back/front.) Our hypothesis here is that there is something in the child’s perception of the relationship such that direct interactions with the caretaker are avoided.

6. This sign is scored if the child only responds to Task VII, (Only VII), where the doggie is having dreams of Mom and then Dad. We find this sign solely in termination-of-parental-rights cases. It is as if the child does not want to think about (and hence react to) anything whatsoever about his or her relationship with either parent. The only thing such a child will permit him- or herself to think about is a dream-for-a-better-future. A typical response is: “The doggie is with his Mom/Dad, and they have a good time.” No responses are offered to the other six tasks. Since this sign appears fairly frequently in agency cases (12 percent), it has been suggested that a child who responds to only one item may be showing wariness and anger toward authority figures, that is, the mental health professionals at the agency. This may be true, but it is also true that we find this same rate in non-agency cases and we still have to wonder why Task VII, in which children interpret the word “dream” as an invitation to wish for a better future, is the sole task to which they allow themselves to respond. If giving only one response represents anger at the evaluator, why does this behavior not occur in relation to other PORT tasks?

7. This sign is scored when a human figure is drawn with the insides of the body omitted, (No insides). The child draws a shell of body and the inside area is blank. We suspected that this would be clinically significant since it points to the presence of massive repression. It is a visual analog of the child’s wish to not experience any feeling or emotion within the body. It only occurs when the relationship with the access-code caretaker has been exceedingly negative. In adults, it occurs with high frequency among individuals suffering psychotic or borderline conditions.

8. This is scored when the child draws tiny figures which are less than ¼ of an inch high in response to Task I and the figures are placed against the margin of the paper. Our clinical hypothesis matches what we have seen, that this occurs only in children who are frightened of the access-code figure, (Tiny/margin).

9. Task VII shows a picture of a doggie asleep in a bed and the child is told the doggie is having a dream about Mommy and, next, about Daddy. The child is asked to tell each dream. The sign is scored when violence or death is mentioned in the child’s response. It is also scored when the doggie to wants to commit suicide. This sign is called Violence VII.

10. This sign is called “severe parts not integrated.” The body parts are drawn so that they have no contact among themselves, for example, the head detached from the body. Our hypothesis: this occurs when the child is not able to integrate emotional information with cognitive information, (Parts not int.[s]).

11. When a child scribbles over something drawn, there is a conflict associated with the access-code. This sign, called “Crosses out,” applies regardless of what explanation the child gives. A child drew himself and his family swimming in a pool. He drew his parents, and then hoping to represent water, scribbled all over the drawing. This child explained that what was drawn was water. Our clinical observation is that a respondent’s motor responses are always more important than how they are verbally explained. We find this sign where there is a rather serious conflict between the respondent and the caretakers within some specific family system, for example, the family: all together in the same place; trying to cooperate; and so forth. In the present case, the boy’s father was in jail and his mother was a significant drug abuser. Our clinical interpretation was that this child is hurt and conflicted over an inability of his parents to share happy times with him. We have also found this sign, along with the characteristic attempt to “explain away” what is essentially a negation or barrier (see item number 18, “Places barriers”), in cases of substantiated abuse. For example, one child, when asked to draw his family, after drawing the family, drew his bed in the upper left hand corner of the page. He placed this bed within a serious “barrier,” thick lines that completely enclosed it. He explained-away the barrier as a “bunk” bed. One of his caretakers (an uncle) was arrested in the near future for the continual sexual abuse of the boy in the boy’s bed. Significantly, signs like this often appear only when there has once been a decent relationship between the child and the abuser. In our experience, the less there has been an emotional bond that has been violated by abuse, the less current problems in that relationship are likely to show up on the PORT. Had this child been molested by a stranger, the PORT manifestation may not have appeared. (Note well: drawing data should never be considered silver-bullets of sexual abuse.)

12. When the child draws an X-mark over any body part, (with the exception of the whole body, covered above), we have another serious sign. It is often an X-mark over the mouth. It has been found almost exclusively in alienation or intimidation cases, where the alienating or denigrating parent has literally ordered the child not to speak to him or her about the other parent. It typically shows up on Task II, where the child is asked to draw him- or herself in the psychological presence of each of the parents. If the alienating parent is, for example, the father, the X-mark will show up on the child’s mouth in reference to the father, but not the mother (X-mark, body part).

13. This is scored when the parts of a drawing are not properly integrated. The standard manifestation of the sign, called parts not integrated, is when a child draws the arms connected to the head rather than to the torso. This is another sign both theoretically and clinically of discomfort in the psychological presence of the access-code figure (Parts not int.)

14. Asymmetric parts occurs when one body part is twice the size of another. Asymmetries are hypothesized to indicate confusion. We find it almost exclusively where a once good parent has become either permanently, or at least temporarily, a poor parent, leaving the child highly perplexed. (A chronically poor parent is probably not confusing to a child—hurtful maybe, but not confusing.) A typical example is when one parent, say the mother, is terrified about something in reference to the father, although there is no strong evidence that the father is doing anything negative. However, this mother’s terror will affect the child. One manifestation occurred where the mother believed her own mother’s allegation (the maternal grandmother) that the father had committed sexual abuse with the child. This could not be substantiated, in spite of thorough investigations. The mother herself later came to disbelieve the allegation. Nevertheless, at the time the allegations were made, the mother believed them, and basically kidnapped the child, keeping her away from the father for a sustained length of time. The father had been the child’s primary caretaker, since the mother had a job that kept her away from the home. Nevertheless, she had been a caring and decent mother. This child was given the PORT more than four times. (We have conducted several studies to see if the PORT parent-of-choice varies according to which caretaker brings the child to the testing session. The POC rarely changes, and when it does may not reflect an error of measurement, Bricklin & Halbert, ink press-b. However, for perceived procedural fairness, we insist that both caretakers bring the to-be-tested child to the session.) In this case, both parents brought the child to the initial session; on the other three occasions the mother brought the child in for testing. The PORTs were given over a four month period. While the POC did not change (the father was the POC in each instance), every single time the child drew her mother, and this amounted to several dozen instances, there was some Asymmetric part on the drawing of the mother, usually the eyes. There were absolutely no asymmetries in any of the other persons portrayed, including her many older siblings and her father. If we take the visual metaphor seriously, that is, that a drawing in some way is a visual analog of some emotional state having to do with secure and safe distances, (explained later), the child is saying “I cannot integrate something I know about my mother to be true (in this case, our prior good relationship), with what I now see to be happening.”

15. This is scored when the mouth or eyes are omitted. Our hypothesis: this occurs when the child, usually blatantly, has been ordered either not to “see something” about a parent or where the child cannot say what he or she wants to say with comfort, (Mouth/eyes omitted).

16. This sign is called “Floaters.” The child draws his or her self-representation on a different plane (higher) than where the caretaker is represented. It suggests “I no longer can occupy the same space as the access-code caretaker.” It occurs on Tasks II, IV or V. The child seems to float above everybody else in the drawing. The clinical meaning: “My parents are too preoccupied to be there for me. I must be strong and rise above the situation.” (As mentioned in the preamble to this section, those who have no faith in drawings will reject this reasoning, saying the child does not have the cognitive skills necessary to harbor a phrase such as: “I must rise above the situation…” in reaction to the PORT stimulus. We will offer clinical, empirical and theoretical data to show that even though the child does not have the words, he or she, from somewhere between 2 ½ years to 3 years of age, does have the necessary genetic structure to formulate and express such concepts.)

17. This sign is called “In front of.” The child draws him- or herself in front of a represented caretaker. The clinical meaning is similar to the above. We find this in mature-for-their-age children who have learned to stay out of parental conflicts. They are “junior league diplomats” (Bricklin & Elliot, 2000, p. 506). They need to parent the parents, because the parents have “disappeared” into their battles. They put themselves in front of the parents because they become more leaders than followers. They are usually the ones who propose peaceful solutions to family problems. We also frequently find this sign in the PORTs of healthy children who have psychotic or borderline parents. The meaning of the sign would be the same as for those children whose parents have psychologically disappeared for other reasons, that is, the child must “lead” the impaired.

18. Places barriers occurs when the child draws any kind of line or object that physically separates him or her from a portrayed caretaker. It does not matter what the barrier is called. In substantiated sex abuse cases, we frequently find instances where a child will draw his or her hair so that it goes from the head right down to the ground while the abusing caretaker stands to the side. The child will label this “hair,” but psychologically it is a barrier. An example: the father was basically a good father who did many fine things with the child, (thirteen years of age) when he, the father, was alone with the child. But this father hated the mother. The child was terrified of the father’s hatred of the mother. In virtually every family drawing that the child rendered, he managed always to place something between himself and the portrayed father. We have seen this in countless cases where there is some fear factor between the child and the portrayed caretaker. Interestingly, when observing this boy with his father, the facial expression he most frequently manifested was one described by Ekman as a fear-smile, a blend of both emotions (1985, pp. 151-152). The conflict revealed by this sign can range from serious to mild, so it cannot differentiate serious situations from those less serious. It does not, in our data, differentiate sexual abuse cases from a case like the one described.

19. This is scored when the dog on Task VII is either “lost” or “cannot find his home.” The child is confused about what he or she can consider “home,” (Lost VII).

20. The child-figure is drawn much bigger than the access-code figure. The meaning is similar to the meaning of PORT signs 16 and 17, (Bigger/parent).

21. The child embellishes the stick figure on Task II or III by drawing hair on the figures, putting eyes on the figures, or adding clothes to the figure. We believed this sign would be typical in high conflict cases where a parent who was once available to a child is no longer so. The embellishment is the child’s attempt to say this parent is no longer what he or she used to be and is trying to “recapture” the parent once enjoyed, (Embellishes).

22. This is scored when any figure is drawn with a moderate slant, about fifteen degrees. Prior research showed that a forty-five degree slant almost always occurs under negative circumstances, (Slant).

23. “Evidence of alienation,” (Alienation) occurs when a child places him- or herself closer to Parent A on Tasks I and II but represents a tie on Task III. Task I has the child draw each parent. Task II has the child in the situation of being alone with each of the parents. Task III represents the child with both parents at the same time. The hypothesis was predicted to occur when a child indeed wants to be with Parent B but only when Parent A is there to provide protection or when the child is fearful of allowing Parent B to see any kind of obvious affection manifested toward Parent A. (If proven, this sign, by itself cannot differentiate these two situations.)

Results

Table 2 presents the probability values for the 23 PORT signs under investigation in each of the four groups.

(Insert Table 2 about here.)

Since these data result from the enumeration of occurrences (which yields a non-continuous ordinal scale), the appropriate statistic to test the null hypothesis would be a chi square. But because of the presence of cells with a frequency of zero or one, the Fisher’s Exact Test, based on the hypergeometric series, yields more accurate probabilities. In the Discussion section, more information will be presented that the PORT signs are best conceptualized as discrete, negative “quality of life” signs that reflect highly disparate clinical situations.

Seventeen of the 23 signs achieved a p of .05 or less that their differential occurrences among the groups could be random. Inspection also suggests there is little difference in the occurrence of the signs among the children in Groups I and II. Effect size estimates are given in Table 3. Eighteen of the 23 effect size values are greater than .30, seven are larger than 0.50, and four exceed 0.75. Taken as a whole, or even evaluated individually, the results suggest that a random process would not yield these figures.

Discussion of Results

Whether there be one or many of the signs, there is an empirical basis to launch an expanded evaluation. Red-flags, though not dispositive, are helpful. Think of an elevated body temperature; it is far from conclusive, but is almost always a sign to launch an expanded medical evaluation. One could also argue that one of more signs suggest a need for psychotherapy among the family members. (WARNING: traditional psychotherapy with high conflict caretakers rarely helps. More frequently it makes things worse. See Bricklin & Elliot, 2000, on the techniques needed in high conflict cases.)

Clinical information as well as the occurrences listed in Table 2, suggest that these signs are best seen as discrete, negative quality-of-life signs, which while all negative, are tremendously different in clinical implications. Because of these disparities, they are not meaningfully additive, even though the cases in Groups III and IV produced more of them than those in Groups I and II. For example, the signs indicating depression and self-directed rage (e.g., signs 3, 6 and 9) are not only much more serious than those indicating, for example, a child’s disappointment over a parent’s post-divorce behavior toward the other parent (signs 11 and 19), but inherently limit the number of negative signs a child is likely to produce. Depression greatly curtails interest in one’s interpersonal environment, the very thing a willingness to respond to the PORT tasks requires. Many of the most seriously disturbed children in Group IV produced only one negative sign, further evidence that the signs are not meaningfully additive.

Another issue involves what the PORT measures. Some PORT seminar attendees say: “This test only measures a child’s perceptions. What if a child perceives a competent parent as incompetent?” But note that validation procedures used independent expert assessments of the quality of the interactions between child and caretaker. Since validity data run in the 90 percent range, one can conclude that the PORT’s manifestations extend to far more than a child’s perceptions. Further, a child’s “perception” played no role in the formation of the four experimental groups. Hence it makes no sense to criticize PORT (or BPS) data because they represent “only” a child’s perceptions. Such a criticism denotes a misunderstanding of what the concept “validity” actually means in science. The critical question is: “Regardless of what a measurement tool happens to be measuring, what can its measurements predict?” The answer to this question is to be found in the details of the validity information.

Translating PORT Data Into Accurate and Relevant Clinical Hypotheses

A conclusion from any single data-source should be considered a hypothesis to be further explored via independent, convergent lines of evidence. I am not aware of any drawing manifestations that point to only one possible conclusion, and there are none that are either necessary and/or sufficient to establish any particular hypothesis.

In weighing the value of any test in use, if we are to avoid the kind of all-or-none review articles we have recently witnessed regarding the Rorschach test and the replies to them (Exner, 2001, pp. 386-388; 2002, pp. 391-404; Ganellan, 2001; Garb, Wood, Lilienfeld, & Nezworski, 2002, pp. 455-457; Meyer, 2001, pp. 389-396; Weiner, Spielberger, & Abeles, 2002, pp. 7-12; Wood, Nezwarski, Garb, & Lilienfeld, 2001, pp. 350-373) and those that involve basic disagreements about what the word “evidence” means in the phrase “evidence-based” (formerly, “empirically-validated”) practice (Anthony, Rogers & Farkas, 2003; Gonzales, Ringeisen & Chambers, 2002) researchers must provide more highly explicated lines of reasoning that link evidence to conclusions than psychometric indices provide. It must include a detailed description of one’s predictor and criterion concepts, especially the empirical equivalents of the concepts. These are the totally value-driven choices of what sensory experiences exemplify a concept (Bricklin & Bricklin, 1999; Bricklin & Halbert, in press-a, in press-b). One must consider whether the measurement scales used by a test are adequate for their (predictive) purposes (measurement theory issues). In our case, this involves specifying when ordinal data are sufficient for a given purpose and those instances where interval data are needed. One must consider whether an appropriate reference standard (group, single-participant, criterion) has been employed, especially when dealing with family systems. The value of measurement data to a decision-maker should be considered. This involves how the decision-maker intends to use new information, what is already believed about the decision area, the strength of these beliefs, what other information is available, the probability that choices will be made with and without the new information, the time-urgency of the needed decision, and the cost (risk level) of making less than an optimum choice (Bricklin & Halbert, in press-a). While psychometric data are needed for such an assessment, they, by themselves, are vastly insufficient to establish such value. Even tools that have low (but known) accuracy rates may still have considerable value in certain situations. When critics of a test have a dispute, they may all be right, since they could be making different assumptions about the above-listed areas.

In the sections to follow, four areas that impact the effective use of drawing data will be described.

The Need For Precise Access-Codes

Drawings are motor-expressed associations of subjective interpretations of verbal instructions. The less the instructions trigger associations to a specific entity, the more difficult it becomes to know what they might predict. We have found, for example, huge differences in how children respond to the task of drawing a specific parent as opposed to a request to draw a “person.” One attendee at a conference sneeringly remarked that he could not understand where drawing research was going. He said: “The next thing you know we’re going to be asking people to draw their Aunt Tilly.” Quite so. If you want someone’s associations to Aunt Tilly that is precisely who you would ask them to draw. You will not get associations to Aunt Tilly if you ask someone to draw a tree, unless Aunt Tilly happens to look (and perhaps act) like a tree. It is not impossible to identify predictor and criterion variables in drawings of houses and trees, but the less the certainty about what such stimuli access mentally, the more difficult it becomes to formulate highly specific validity criteria.

In our experience whatever else drawings can predict, they do best at portraying motor-driven associations to a specific stimulus. The more vaguely known the stimulus is, the harder it is to determine what the testee’s motor-driven associations are in response to. “Aunt Tilly” is more specific than a “person,” “Draw your house” is more specific than “Draw a house,” and “Draw your Mother” is more specific than “Draw a female.”

What Can Drawing Data Accurately Predict?

We believe drawings can be useful, but within a narrow range of endeavors. Our experience (more than 40 years, often with much collateral and follow-up information) suggests drawings do best at predicting basic, often unconscious, quickly occurring motor dispositions, for example, to move toward, away from, or against a precisely-known access-code. Empirical information on the observational protocols that suggest this is the case can be found in Bricklin & Elliot (2002a; 2002b) and in Bricklin & Halbert (in press-a, in press-b). It is not surprising that the most “genuine” things about a person should be manifested mainly in non-verbal media (Ekman, 1985; Ekman & O’Sullivan, 1991). Further, the most sincere “truths” about an individual do not necessarily last very long even in their non-verbal expressions. Some spontaneous facial expressions last less than one second and require special equipment to observe (Bricklin & Elliot, 1998).

A critical issue in using drawing data to make accurate assertions is the role of bias. We encounter at our training seminars widespread misunderstanding of the concept of bias, for example, that one needs to “adjust” for poor visual-motor skills or “drawing ability” when using drawing data. The need for “adjustment” does not follow automatically from the operation of some variable, regardless of whether it co-creates the measurement response. One needs to correct for bias only when a variable creates bias in regard to a particular predictive usage. (Would you like to go to a surgeon who made it through medical school because some professor said, “Let’s raise this kid’s surgical-success scores—we should make an adjustment for his miserable perceptual-motor skills.”)

Empirical data from the PORT strongly suggest that few of the variables thought to create bias in using drawings to make defensible assertions actually do so. Validity figures, largely in the 90 percent range, have been obtained in 1,508 cases—cases which represent a hugely diversified group in age (3 ½ years to 16 years, with some data from 2 ½ year-olds) intelligence, age-appropriate (and non-appropriate) school grade placements, visual-motor skills, presence and absence of learning disabilities and/or ADHD diagnoses, socioeconomic levels and drawing ability (Bricklin & Elliot 2002a, 2002b; Bricklin & Halbert, in press-b).

What about “drawing ability”? First, it is clear from the thousands of PORTs reviewed, that some children render more “artistic” drawings than others. Our premise is that one can extract many useful predictive indices from drawings unrelated to artistic abilities. A much more basic force than “art ability” drives drawings. At the frontiers of science is the awareness that the universe (and whatever one can say scientifically about anything in it) must arise from a paradigm in which the description of any entity within the system has to be made in reference to other things that are part of this system. Consequently “… (I)f something has a position, that position can only be defined with respect to… other things in the universe” (Smolin, 2001, p. 18). If something is in motion, one can only become aware of and describe it in terms of its unfolding relationships with other entities. Space is the relationship among things in the real world. “Time” has no absolute meaning; it literally is the change of some entity in relation to another. These same concepts hold in terms of up-to-date attempts to reconcile Einstein’s ideas with those from quantum mechanics. One of the most advanced theories, called “loop quantum gravity,” is seem as a “structure of relations…” (Smolin, 2001, p. 138; Peat, 2002, pp. 15-17). Indeed mathematics is the “science of patterns” (Devlin, 2000, p. 7). Cutting-edge thinkers believe, along with Smolin, that the “… world is nothing but an evolving network of relationships” (p. 20). Nobel prize winner Daniel Kaneman understands human perception this way (2003, p. 703).

All of this—and common sense—suggest that one of the most basic things an organism is pre-wired to know is how it relates to other discrete entities. These relationships not only underly most of what psychologists are interested in about people, but determine what humans come to experience as space and time. A basic survival skill an organism must have is the ability to rapidly discern how it relates to other entities, human and otherwise, and with this the ability to know when to move toward, away from or against them. The rapid approximation of safe and unsafe distances would seem to be a cosmic feature of the universe and a necessary tool for survival (Cozolino, 2002, p. 235). We believe humans make these calculations constantly, and that drawings provide an ideal medium in which to express them, since their mental representations depend primarily on visual/spatial and spatial/kinesthetic information.

The next issue is to consider whether children have the means to complex relationships such as “I must rise above this painful situation.” We can look at what linguists call the “deep structure” or the pre-wired network that allows humans, starting at a very early age, to comprehend the world (Chomsky, 1957; 1968). Children apparently have the ability to comprehend important parts of the world long before they have the words to express them. Deacon claims children have the necessary mental structure by 18 months (though not necessarily the words), to comprehend and motorically express complex relationships among patterns (Deacon, 1997, pp. 6, 24, 26, 37, 207). This same point is made by Calvin and Bickerton (2001). They, however, believe the necessary structure is in place by 36 months rather than 18 months (p. 163).

Since our overall argument is that spontaneous motor expressions can represent meaningful communications about relationships among discrete entities even from very young children who lack the words to express these relationships, we consider what the evolutionary purpose of the deep structure is thought to represent. Current thinking is that the evolutionary purpose was not to provide for an increased ability for humans to speak to one another, but rather for humans to more adequately represent their worlds and the relationships in them (Calvin & Bickerton, 2001). This is a key point, as it puts the spotlight on the development of neuropsychological processes that lead to syntactic structures and words as more serving the survival mechanism of an adequate representation of what is “out there” and how one can relate to what is “out there,” than just the ability to speak.

Even if we agree that young humans can process and represent patterns of relations in the real world long before they can speak about them, we might ask where drawings fit into these processes. And where is further evidence that differences in drawing ability need not necessarily require “adjustment”?

Three pieces of evidence will be offered. One posits that even fairly impaired children are capable of executing relatively precise sequences of motor behaviors at ability levels beyond their measured cognitive skills. Next, we will offer data to suggest that many “good” drawings that appear to be rendered spontaneously are in fact not. Professional artists who seem so effortlessly to render “good” spontaneous drawings are actually doing something else. Finally, we will offer evidence that the ability and even the very urge to draw were created by evolutionary forces aimed at helping humans relate safely to their caretakers.

Caretakers of mentally retarded children have noted that these children can successfully execute rather complex motor sequences at efficiency levels far beyond their general intellectual capacities (Calvin & Bickerton, 2001, p. 37). Hence it would seem that the ability levels required to express relationships among entities motorically are less than those needed to talk about these relationships.

Next, consider the fact that some people can not only draw more artistic creations than others, but further, can do so in ways that appear so easy and spontaneous that the illusion is created that drawing ability must be controlled for if we are to use them to generate empirically supported clinical hypotheses. Deacon’s data suggest that a “good artist,” in addition to any genetic advantage possessed, can bring a huge, pre-programmed, already-integrated series of motor-expressable sequences to a drawing task that can be executed quickly and only seemingly as though they follow from innate sources (Deacon, 1997, p. 237). Motor sequences “… can be ‘off loaded’ to other motor systems that allow them to be (executed) … as an unanalyzed modular program” (Deacon, 1997, p. 237). This means that highly skilled learned behaviors can emerge as rapidly and seemingly as easily performed as innate behaviors. Many cognitively thought-out pre-wired pieces can be embedded in what otherwise are viewed as “spontaneous” sequences. “Drawing ability,” as the term is commonly used, refers to the degree to which a drawing appears to be reasonably “artistic.” Since the predictor variables are called “drawings,” the ability to “draw” (respond motorically) is confused with the ability to draw artistically.

Our main empirical argument against this assumption is that the validity of the PORT signs are unaffected by the many variables that make some drawings look better than others. Our theoretical argument involves the fact that even very young children can mentally process (“understand”) psychologically important interpersonal relationships, in particular, the degree to which they are safe, in ways that are beyond their ability to express these understandings either in words or “artistic” drawings. We conclude that “good” drawing ability is not needed for what the PORT seeks to explicate. There is the further interesting point—though not crucial to our empirical argument—that the artistically superior spontaneous drawings rendered by some persons are not really being driven by occurring in-the-moment (spontaneous) psychological forces.

The final issue concerning what drawings can predict centers on a theory by Dissanayake (2002). She believes the evolutionary principle behind artistic impulses is psychological in nature. The psychology/art area is replete with theories that relate psychological functions to artistic themes. Art texts view early cave (and other primitive) drawings as representing magical ways of “controlling” the world or as having magical/historical intentions (Egyptian art). Other theories are even more specifically psychological. Robert Waelder (1965) sees drawings as human attempts to resolve tension (p. 6), satisfy frustrated wishes (p. 24) and as a general ego “problem-solver” (p. 44). The esteemed historian of art, Heinrich Wolfflin, (1931) covered these themes by addressing surface proportions (p. 25), and the relationships of parts to wholes (pp. 19-149). For our purposes, this theme is more directly covered in the work of Dissanayake. Her belief is that artistic impulses represent the child’s need to develop a safe and effective accommodation with his or her caretakers. The theme is also reflected in the research on evaluative responding, that is, responses that are immediate, consciously unintentional as well as implicit, and automatically wired to motor predispositions to move “toward or away from” (Duckworth, Bargh, Garcia & Chaiken, 2002, pp. 513-519).

One final point can be made about what drawings can predict. It has been our experience, as well as that of Zygmunt A. Piotrowski (see Piotrowski & Biele, 1986) that regardless of what one is asked to draw (e.g., a tree, your uncle, a building, your spouse) one always (primarily) draws oneself as one relates to the access-code. If a person is asked to “draw your uncle” and the respondent draws an “angry” uncle, it may be that this uncle is indeed perceived as angry (what Piotrowski calls a “secondary meaning”). But what it means for sure is that the respondent is angry at his or her uncle, that is, that the access-code elicits associations of anger in the respondent. This anger may be in response to the uncle’s anger—but this may not be the case. It may be something else about the uncle that is mobilizing the anger. What can be concluded with a much higher degree of certainty is that it is the repsondent who is angry at the uncle, while it is less clear as to what it is about the uncle that is eliciting the anger.

It is perhaps our bilateral symmetry, our need to calculate safe and unsafe distances to all things in our environments and our continuing relationship to gravity (see the works of Moshe Feldenkrais) that account for the fact that whatever we draw, we are in some ways drawing ourselves. Many art historians seemed puzzled by how much the Mona Lisa looks like her creator (DaVinci), but this may be hard-wired into the drawing process itself and not an accident or joke.

Where Can One Find The Kind Of Data Needed To Validate What Drawings Can Predict?

Many of the most critical things psychologists have to predict about people are known only to their therapists or family therapists—professionals who worked with them over long time spans and were able to gather data from multiple and independent sources (Miller, 1998; Piotrowski & Biele, 1986). This is also true of custody evaluators, who spend several months gathering data and have multiple and independent information from tests, interviews, collateral informants and especially real-life observations. Examples of such (case-specific) criterion are given in Bricklin & Halbert, (in press-b). When such data are lacking, highly operationalized empirical observation protocols are needed.

Systems Complexities in Interpersonal Predictions

Mental health professionals often think of a system as an interactional model in which stable traits of person A interact with stable traits of person B. One sees this in the way they conceptualize and write about their custody (and other psychological) evaluations. There are sections called “Mister Jones,” “Mrs. Jones,” child “Mary Jones,” child “Sam Jones,” as though one can assess each element in a system as a separate entity and then somehow add up the parts. In systems-based decisions, the elements of the system cannot be evaluated apart from their interactions within the system. A system is composed of two or more elements, in which the following conditions also apply. Each element has an effect on the whole system. The elements (and their effects) are not only interdependent but however subgroups are formed, none can have an independent effect on the system as a whole. One cannot influence a system of which one is a part, and then not oneself be impacted by this influence (Ackoff, 1999, pp. 15-17). According to Ackoff, people used to mechanistic models of science are often confused by systems thinking. The former approaches understanding and/or prediction via a deconstruction process; elements are analyzed one by one, after which their interactions with other elements are analyzed. With systems, the elements can only be analyzed as they operate within the system. In the field of psychology, predictions often need to be system-specific. When this is the case, predictor-models must be able to discern and specify the interpersonal systems in which a prediction is likely to be true. While there are predictive models that attempt to yield assertions that are not system-specific (principles of symmetry or invariance, Weinberg, 1994, pp. 136-137), within psychology, these predictions are usually vague. They are hard to validate by means of specifying exact events as they will occur in the real world.

We will confine our discussion here to the effects systems have on the attempt to validate the relationship between the degree-of-closeness between a child and caretaker as observed on a drawing, and the degree-of-closeness likely to be observed in some real-life context. Child custody situations furnish an ideal medium to reveal the importance of systems complexity, since a disputed post-divorce world is a nightmare of suddenly shifting allegiances, often within a framework where a “safe choice” for a child involves only two possibilities. To make even semi-accurate predictions in such a chaotic world requires a sharpened awareness of the system-specific nature of prognosticated behaviors. Let us offer the following example.

Research, already cited, indicate that each PORT task represents the child’s reactions to a different family system. Task V, for example, demonstrates a very robust relationship between the parent-of-choice on that item, and the parent seen by independent experts as the caretaker who most supports family unity. The following case illustrates our concerns. The father was poorly attuned to the day-to-day parenting of his six year old child. He loved him, but took little interest in his daily life. After the divorce, he quickly remarried. Father, however, realized soon after the divorce what he had lost in his opportunity to remain close to his child on a day-by-day basis. He consequently actively and sincerely promoted family unity and cooperation. Mother, hurt and angered by his quick remarriage, at least temporarily did not. The PORT, administered around this time, showed the mother to be the parent-of-choice on every PORT task (I, II, III, IV, VI and VII), which all clinical data supported, yet the father was the choice on Task V, the task which reflects the child’s experience of the degree to which each parent supports family unity. Critics of the “closeness” principle would see this as evidence that it is not valid, since its predictions are inconsistent. Our empirical and clinical data strongly suggest that the closeness of a self-representation to an access-code on a drawing indeed can find confirmation in the external (validating) world, but only if it is recognized that closeness-to-a-caretaker may be system-specific.

Four Caveats in the Clinical Use of Drawings

These four complexities make certain uses to which drawings have been put precarious endeavors, for example, the attempt to use drawings to diagnose sexual abuse. The first complexity enters the picture since an evaluator can hardly formulate a precise access-code for sex abuse cases: “Draw a picture of Daddy touching your body.” Any response to such an access-code would be tainted by its leading nature. And as one moves to less precise access-codes, for example, “Draw a picture of: you and Daddy, what you did some certain day, what hurts you, embarrasses you” and so forth, the less one can be sure what a respondent is associating to. The second complexity (what a drawing manifestation is able to predict) arises with the fact that drawings (and most psychological tests) do best at revealing what a person has experienced, and not the specific events that created that experience (absent a good bit of construct-validity evidence, plus the use of precise access-codes). Further, many abusers “sell” their behaviors to children under non-threatening frames of reference (“This is Daddy’s tickle-game”) so nothing much registers that can be detected by tests that aim to discover negative psychological events. Further, the attempt to identify the cause of any diagnostic manifestation is always difficult. A radiologic film can reveal that “something” is present, but what that “something” is and especially what forces co-caused it to be there are more difficult to determine. The third complexity, the problem of finding accurate validating evidence, is quite vexing in the sex abuse area, since there are so few ways to prove abuse ever occurred absent crystal-clear physical evidence and/or meaningful confession. Finally, the fourth difficulty, the role of systems complexities, enters the picture. There may be many contexts in which a child and an abusing parent have a comfortable relation. I observed a 4 ½ year old girl sitting lovingly in her father’s lap in my waiting room. She was responding warmly to his kisses, holding him and he was behaving appropriately with warmth and attentiveness. A few months later there was conclusive evidence (meaningful confession and physical data) that sexual abuse had begun at least one year prior to my observations. As so called “states of arousal” shift (Bricklin, 1995, pp. 96-100) so too do available psychological resources, often radically. Successful predictions in situations which strongly raise the likelihood of shifts in states of arousal require systems concepts. (Interestingly, her fear of her father did show up in her PORT.)

These four complexities reflect our caveats about using drawing manifestations in assessment. Researchers must create precise access-codes directly related to what they seek to predict and there should be a reasonable expectation that drawing manifestations are able to predict them. The more that verbal and highly abstract elements are required for the respondent to understand the stimulus instructions, the less likely is it that predictive endeavors will succeed since there is no way to determine what the subject is associating to. Third, the researchers need to locate unbiased and accurate validity sources. Forty years worth of data suggest that drawings do best at predicting rapidly occurring motor movements in system-specific contexts, having mostly to do with the rapid calculations of safe versus unsafe distances, that is, when to move toward, away from or against the access-code. This applies even when someone renders a drawing where no instructions were given, only here there is no way to guess who or what the access-code is. Fourth, the relation between systems complexity and accurate predictions needs to be understood.

References

Ackoff, R. L. (1999). Ackoff's best. New York, NY: Wiley.

Anthony, W., Rogers, E. S., & Farkas, M. (2003). Research on evidence-based practices: Future directions in an era of recovery. Community Mental Health Journal, 39 (2), 101-114.

Bargh, J.A. (1997). The automaticity of everyday life. In R.S. Wyer, Jr. (Ed), The automaticity of everyday life: Advances in social cognition (Vol. 10, pp. 1-61). Mahwah, NJ: Erlbaum.

Bricklin, B. (1984). The Bricklin perceptual scales: Child-perception-of-parents series. Furlong, PA: Village Publishing, Inc.

Bricklin, B. (1989). The perception-of-relationships test. Furlong, PA: Village Publishing, Inc.

Bricklin, B. (1992). Data based tests in custody evaluations. The American Journal of Family Therapy, 20(3), 254-265.

Bricklin, B. (1995). The custody evaluation handbook: Research-based solutions and applications. New York: Brunner/Mazel.

Bricklin, B. (1997). Test manuals supplement update. Furlong, PA: Village Publishing, Inc.

Bricklin, B. (1998). Sequence of steps and critical assessment targets of a comprehensive custody evaluation. In G.P. Koocher, J.C. Norcross, & S.S. Hill III (Eds.), Psychologists’ desk reference (pp. 499-502). New York, NY: Oxford University Press.

Bricklin, B. (1999). The contribution of psychological tests to custody-relevant evaluations. In R.M. Galatzer-Levy & L. Kraus (Eds.), The scientific basis of child custody decisions (pp. 120-156). New York, NY: Wiley.

Bricklin, B., & Bricklin, P. M. (1999). Custody data as decision-theory information: Evaluating a psychological contribution by its value to a decision-maker. Clinical Psychology: Science and Practice, 6(3), 339-343.

Bricklin, B., & Elliot, G. (1995). ACCESS: A comprehensive custody evaluation standard system. Furlong, PA: Village Publishing, Inc.

Bricklin, B., & Elliot, G. (1997). Test manuals supplement. Furlong, PA: Village Publishing, Inc.

Bricklin, B., & Elliot, G. (1998, August). Therapeutic and monitoring strategies in high conflict custody and visitation disputes using new motor therapy strategies to increase cooperation. Paper presented at the American Psychological Association, San Francisco, CA.

Bricklin, B., & Elliot, G. (2000). Qualifications of and techniques to be used by judges, attorneys and mental health professionals who deal with children in high conflict divorce cases. University of Arkansas At Little Rock Law Review, 122(3), 501-528.

Bricklin, B., & Elliot, G. (2002a, April). What can empirical data on 4,500 child custody cases tell us? Paper presented at the American College of Forensic Psychology, San Francisco, CA.

Bricklin, B., & Elliot, G. (2002b). The Perception-of-Relationships Test (PORT) and Bricklin Perceptual Scales (BPS): Current and new empirical data on 3,880 cases, 1961-2002. Furlong, PA: Village Publishing, Inc.

Bricklin, B., & Elliot, G. (in press-a). Empirically-assisted family assessment. In L. Gunsberg & P. Hymowitz (Eds.), Divorce and custody: Contemporary developmental psychoanalytic perspectives. New York, NY: The Analytic Press.

Bricklin, B., & Elliot, G. (in press-b). Test-assisted detection of the Parental Alienation Syndrome. In R. Gardner & R. Sauber (Eds.), International Handbook of Parental Alienation Syndrome: Clinical and Legal Issues. New York, NY: Haworth Press.

Bricklin, B., & Halbert, M. (in press-a). Can child custody data be generated scientifically? The American Journal of Family Therapy.

Bricklin, B., & Halbert, M. (in press-b). Perception-of-Relationships Test and Bricklin Perceptual Scales: Validity and Reliability. The American Journal of Family Therapy.

Calvin, W. H., & Bickerton, D. (2001). Lingua ex Machina: Reconciling Darwin and Chomsky with the human brain. Cambridge, MA: MIT Press.

Chomsky, N. (1957). Syntactic structure. The Hague, Netherlands: Monitor Press.

Chomsky, N. (1968). Language and mind. New York, NY: Harcourt, Brace & Javanovitch.

Cozolino, L. J. (2002). The neuroscience of psychotherapy: Building and rebuilding the human brain. New York, NY: Norton.

Deacon, T. W. (1997). The symbolic species: The co-evolution of language and the brain. New York, NY: Norton

Devlin, K. (2000). The math gene: How mathematical thinking evolved and why numbers are like gossip. London, England: Widenfeld & Nicolson.

Dissanayake E. (2000). Art and intimacy: How the arts began. Seattle, WA: University of Washington Press.

Doolittle, D. B., & Deutsch, R. (1999). Children and high conflict divorce: Theory, research, and intervention. In Galatzer-Levy & L. Kraus (Eds.), The scientific basis of child custody decisions (pp. 425-440). New York, NY: Wiley.

Duckworth, K. L., Bargh, J. A., Garcia, M., & Chaiken, S. (2002). The automatic evaluation of novel stimuli. Psychological Science, 13(6), 513-519.

Ekman, P. (1985). Telling lies. New York, NY: Norton.

Ekman, P., & O'Sullivan, M. (1991). Who can catch a liar? American Psychologist, 46(9), 913-920.

Exner, J. E. (2001). A comment on “The misperception of psychopathology: Problems with the norms of the Comprehensive System for the Rorschach.” Clinical Psychology: Science and Practice, 8 (3), 386-388.

Exner, J. E. (2002). A new nonpatient sample for the Rorschach Comprehensive System: A progress report. Journal of Personality Assessment, 78(3), 391-404.

Ganellen, R. J. (2001). Weighing evidence for the Rorschach’s validity: A response to Wood et al. (1999). Journal of Personality Assessment, 77 (1), 1-15.

Garb, H. N., Wood, J. L., Lilienfeld, S. O., & Nezworski, T. M. (2002). Effective use of projective techniques in clinical practice: Let the data help with selection and interpretation. Professional Psychology: Research and Practice, 33 (5), 454-463.

Gonzales, J. J., Ringeisen, H. L., & Chambers, D. A. (2002). The tangled and thorny path of science to practice: Tensions in interpreting and applying “evidence.” Clinical Psychology: Science and Practice, 9 (2), pp. 204-209.

Hoppe, C. F. (1993, August). Test characteristics of custody-visitation litigants: A data-based description of relationship disorders. In S. Podrygula (Chair), Empirical approaches to child custody determination. Symposium conducted at the meeting of the American Psychological Association, Toronto, Ontario, Canada.

Hoppe, C. F. (1997). Perpetually battling parents. In B. Mark and J. Incoravia (Eds.), The handbook of infant/child and adolescent psychotherapy, Vol. II. (pp. 485-501). Northdale, NJ: Aaronson.

Hoppe, C. F., & Kenney, L. (1994, August). Characteristics of custody litigants: Data from the southern California group. Paper presented at the meeting of the Psychological Association, Los Angeles, CA.

Hoppe, C. F., & Kenney, L. (1995, August). MMPI-2 and Rorschach profiles of custody litigants: An intercorrelational study. Paper presented at the meeting of the American Psychological Association, New York, NY.

Hoppe, C. F., & Kenney, L. (1997, August). Therapeutic intervention in high conflict divorce: Countertransference and the horrible decision. Paper presented at the meeting of the American Psychological Association, Chicago, IL.

Horney, K. (1945). Our inner conflicts. New York, NY: Norton.

Kaneman, D. (2003). A perspective on judgment and choice: Mapping bounded reality. American Psychologist, 58 (9), 697-720.

Kelly, J.B., & Lamb, M.E. (2000). Using child development research to make appropriate custody and access decisions for young children. Family and Conciliation Courts Review, 38 (3), 297-311.

Kerig, P. K., & Lindahl, K. M. (Eds.). (2001). Family observational coding systems: Resources for systemic research. Mahway, NJ: Lawrence Erlbaum Associates.

Lahey, B. B., Conger, R. D., Atkeson, B. M., & Trieber, F. A. (1984). Parenting behavior and emotional status of physically abusive mothers. Journal of Consulting and Clinical Psychology, 52, 1062-1071.

Meyer, G. J. (2001). Evidence to correct misperceptions about Rorschach norms. Clinical Psychology: Science and Practice, 8, 389-396.

Miller, R. B. (1998). Epistemology and psychotherapy data: The unspeakable, unbearable, horrible truth. Clinical Psychology: Science and Practice, 5(2), 242-250.

Piotrowski, Z.A., & Biele, A.M. (1986). Dreams: A key to self knowledge. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.

Schutz, B. M., Dixon, E. B., Lindenberger, J. C., & Ruther, N. J. (1989). Solomon’s sword: A practical guide to conducting child custody evaluations. San Francisco, CA: Jossey-Bass Publishers.

Smolin, L. (2001). Three roads to quantum gravity. New York, NY; Basic Books.

Solomon, J. & Biringen, Z. (2001). Another look at the developmental research: Commentary on Kelly and Lamb’s “Using child development research to make appropriate custody and access decisions for young children.” Family and Conciliation Courts Review, 39 (4), 355-364.

Thomas, G.V., & Jolley, R.P. (1998). Drawing conclusions: A re-examination of empirical and conceptual bases for psychological evaluation of children from their drawings. British Journal of Clinical Psychology, 37, 127-139.

Waelder, R. (1965). Psychoanalytic avenues to art. New York, NY: International Universities Press.

Weinberg, S. (1994). Dreams of a final theory. New York, NY: Vintage Books.

Weiner, I. B., Spielberger, C. D., & Abeles, N. (2002). Scientific psychology and the Rorschach Inkblot Method. The Clinical Psychologist, 55 (4), 7-12.

Wolfflin, H. (1931). The sense of form in art. Munich, Germany: F. Bruckman.

Wood, J. M., Nezworski, T. M., Garb, H. N., & Lilienfeld, S. O. (2001). The misperceptions of psychopathology: Problems with the norms of the Comprehensive System for the Rorschach. Clinical Psychology: Science and Practice, 8 (3), 350-373.

Zajonc, R.B. (1980). Feeling and thinking: Preferences need no inferences. American Psychologist, 35, 151-175.

Zajonc, R.B. (1998). Emotions. In D.T. Gilbert, S.T. Fiske, & G. Lindzey (Eds.), Handbook of social psychology (4thed., Vol. 1, pp. 591-632). New York: Oxford University Press.

Table 1

PORT Task-Difference-Scores (TDS) and Changes in Parent-of-Choice (POC) from Test 1 to Test 2

PORT TDS	Frequency at Test 1	Frequency at Test 2	POC Changed	Percent POC Changed	Percent POC Unchanged with ³TDS
7	1	1	0	0.0
6	4	4	0	0.0	100.0
5	5	6	0	0.0	100.0
4	14	12	1	7.1	95.8
3	18	20	0	0.0	97.6
2	56	54	2	3.6	97.0
1	17	16	3	17.6	94.8
0	12	14	3	25.0	92.9
Total	127	127	9	7.1	92.9

Table 2

The Frequency of Appearance of 23 PORT Signs Among Four Experimental Groups

	PORT Sign	Group I	Group II	Group III	Group IV	Totals	Fisher’s Exact Probability
1.	Two heads	0	0	0	4	4	.0106
2.	Two figures	1	3	12	14	30	.0011
3.	X-mark	0	0	1	6	7	.0020
4.	Vert.-mis.	1	1	3	4	9	.5019
5.	Back / Front	0	0	3	5	8	.0300
6.	Only VII	0	0	0	5	5	.0018
7.	No insides	0	0	0	4	4	.0106
8.	Tiny / Margin	0	0	3	5	8	.0313
9.	Violence VII	0	1	4	5	10	.1216
10.	Parts not int.(s)	0	0	1	7	8	.0010
11.	Crosses out	0	1	7	7	15	.0238
12.	X / body parts	0	1	15	1	17	.0000
13.	Parts not int.	1	2	1	12	16	.0000

Table 2 (continued)

The Frequency of Appearance of 23 PORT Signs Among Four Experimental Groups

	PORT Sign	Group I	Group II	Group III	Group IV	Totals	Fisher’s Exact Probability
14.	Assym. pts.	2	3	12	10	27	.0493
15.	Mouth / eyes om.	0	1	5	5	11	.1243
16.	Floaters	3	5	19	14	41	.0049
17.	In front of	1	1	9	6	17	.0534
18.	Places bar.	1	2	12	11	26	.0047
19.	Lost VII	0	0	3	4	7	.0857
20.	Bigger / Parent	3	3	14	11	31	.0264
21.	Embellishes	0	0	14	2	16	.0000
22.	Slant	1	4	6	6	17	.6248
23.	Alienation	0	0	18	4	22	.0000
	Total	14	28	162	152

Table 3

Effect Size (ES) and phi values for Fisher's Exact Test Probability (FET) ______________________________________________________________________

Sign FET phi ES Sign FET phi ES

______________________________________________________________________

1. 0.0106 0.2104 0.4305 13. 0.0000 0.3703 0.7972

2. 0.0011 0.2795 0.5823 14. 0.0493 0.1508 0.3051

3. 0.0020 0.2627 0.5446 15. 0.1243 0.1053 0.2119

4. 0.5019 0.0004 0.0009 16. 0.0049 0.2358 0.4853

5. 0.0300 0.1717 0.3486 17. 0.0534 0.1472 0.2977

6. 0.0018 0.2657 0.5513 18. 0.0047 0.2371 0.4881

7. 0.0106 0.2104 0.4305 19. 0.0857 0.1249 0.2517

8. 0.0313 0.1700 0.3450 20. 0.0264 0.1768 0.3593

9. 0.1216 0.1065 0.2143 21. 0.0000 0.3703 0.7972

10. 0.0010 0.2821 0.5880 22. 0.6248 0.0290 0.0581

11. 0.0238 0.1808 0.3677 23. 0.0000 0.3703 0.7972

12. 0.0000 0.3703 0.7972