Are Parents Accurate Reporters of Their Child s Cognitive Abilities?

Journal of Psychopathology and Behavioral Assessment, Vol. 22, No. 1, 2000 Are Parents Accurate Reporters of Their Child’s Cognitive Abilities? Danie...
Author: Edwina Gregory
0 downloads 0 Views 76KB Size
Journal of Psychopathology and Behavioral Assessment, Vol. 22, No. 1, 2000

Are Parents Accurate Reporters of Their Child’s Cognitive Abilities? Daniel A. Waschbusch,1,4 Eric Daleiden,2 and Ronald S. Drabman3 Accepted: December 22, 1999

Examined parent’s ability to accurately report their child’s cognitive functioning. Participants were 145 children and their parents referred to an outpatient mental health clinic for cognitive testing. Parent reports were measured using Likert ratings designed for research and clinical purposes. Children’s cognitive abilities were measured using the Woodcock-Johnson Psycho-Educational Battery—Revised (Woodcock, 1989) and the Wechsler Intelligence Scale for Children—Third Edition (Wechsler, 1991). For boys, parental reports of general cognitive ability, fluid reasoning, comprehension-knowledge, visual processing, auditory processing, and acquisition and retrieval were significantly related to performance-based measurements of these same traits, and parental report of boys’ fluid reasoning and visual processing ability evidenced specificity. For girls, parental report of general cognitive ability, fluid reasoning, visual processing, and auditory processing were significantly related to performance-based measurements of these same traits, and parental reports of girls visual processing and auditory processing evidenced specificity. These findings suggest areas where clinicians can be more confident of parental report of children’s cognitive abilities and other areas where clinicians should be wary of parental report. KEY WORDS: parents reports; Woodcock-Johnson psychoeducational battery—revised; Wechsler intelligence scale for children—third edition.

1

Department of Psychology, Dalhousie University, Halifax, Nova Scotia, Canada B3H 4J1. Department of Psychology, University of Tulsa. 3 Department of Psychiatry and Human Behavior, University of Mississippi Medical Center. 4 To whom correspondence should be addressed at Department of Psychology, Dalhousie University, Halifax, Nova Scotia, Canada B3H 4J1 or by internet to [email protected]. 2

61 0882-2689/00/0300-0061$18.00/0  2000 Plenum Publishing Corporation

62

Waschbusch, Daleiden, and Drabman

INTRODUCTION Parent ratings of their child’s behavior are commonly used to assess children’s psychosocial adjustment. Advantages of this methodology, such as low cost, ease of use, and parent knowledge of cross-temporal, crosssetting child behavior, have made parent behavioral ratings among the most popular assessments of children’s adaptive and emotional functioning. A considerable volume of empirical research has supported the validity and diagnostic utility of parent ratings of their child’s behavior as an index of their child’s adjustment (e.g., Achenbach, McConaughy, & Howell, 1987; Fabrega, Ulrich, & Loeber, 1996; Hart, Lahey, Loeber, & Hanson, 1994; Loeber, Green, Lahey, & Stouthamer-Loeber, 1990; Loeber, Green, Lahey, & Stouthamer-Loeber, 1991). Similar studies examining parent ratings of their child’s cognitive functioning are not as readily available. It is likely that this question has not been adequately addressed because cognitive functioning is ideally indexed using performance measures. For example, it is relatively fast and efficient to administer the vocabulary and block design subtests of the Wechsler Intelligence Scale for Children—Third Edition (WISC-III; Wechsler, 1991), and these subtests provide a reasonable estimate of a child’s cognitive functioning (Sattler, 1992). In spite of this fact, examinations of whether parents are accurate reporters of their child’s abilities may be important for a number of reasons. First, parents are often the first to present their child for treatment in clinical and educational settings and therefore provide the initial characterization of a case. Because teacher-reports and testing may be viewed as more costly or invasive procedures in a least restrictive treatment model, some children may not receive objective and intensive cognitive assessment until after parents indicate that their child may have a problem. Second, it is often difficult for parents to describe the nature of their child’s cognitive abilities. Parents, unlike trained professionals, may find it challenging to verbalize even basic concepts of their child’s cognitive abilities, such as the distinction between verbal IQ and performance IQ, without the assistance of a professional. Research examining whether parents can accurately report their child’s cognitive abilities may lead to the development of such vocabulary. For example, there may be specific behavioral referents underlying cognitive abilities that clinicians and teachers could use when communicating to parents (and vice versa) about a child’s cognitive abilities. Finally, many psychiatric intake interviews specifically question parents about their child’s cognitive abilities. Information on the accuracy of this portion of the parent interview might aid clinicians in their evaluations. These assertions suggest that the accuracy of parent reports of their child’s cognitive abilities warrant evaluation.

Parent Report of Cognitive Abilities

63

The purpose of this study was to examine parent reports of their child’s cognitive abilities. To do so, we first selected a theoretical model to provide a framework for characterizing children’s cognitive abilities. We selected Horn-Cattell theory (c.f. Horn, 1985; Horn, 1989; Horn, 1994; Horn & Noll, 1997), which is also known as Gf–Gc theory because of its original distinction between fluid (Gf) and crystallized (Gc) intelligence. HornCattell theory uses multiple factors, referred to as broad cognitive abilities, to explain cognitive functioning. Some of the more important cognitive abilities described in Gf–Gc theory are fluid reasoning, comprehensionknowledge, visual processing, auditory processing, short term acquisition and retrieval, and long-term storage and retrieval. Fluid reasoning refers to the ability to reason under conditions of novelty and includes deductive reasoning and inductive reasoning. Comprehension-knowledge, also known as crystallized intelligence, represents knowledge about a culture including communication comprehension, awareness of conventions, and judgment. Visual processing involves fluency with spatial manipulations and visual stimuli whereas auditory processing involves parsing, apprehending, and anticipating patterns among sound streams. Short term acquisition and retrieval is the ability to encode, retain, and recall information within an immediate situation. Finally, long-term storage and retrieval refers to the breadth and fluency of associative recall of information over extended time periods. Horn (1985) terms these different abilities ‘‘broad cognitive abilities,’’ and he has mapped these abilities to developmental and information-processing hierarchies. That is, fluid reasoning and comprehensionknowledge are adult relational abilities, visual and auditory processing represent youthful perceptual organization abilities, and short-term acquisition and retrieval and long-term storage and retrieval signify childhood associational abilities. We chose the Gf–Gc theory of cognitive abilities because it has received considerable empirical support. A comprehensive review of this evidence is beyond the scope of this paper, but exemplars of the various validational approaches will be presented. Carroll (1993) reanalysed 460 data sets and found evidence supportive of the factorial structure of Gf–Gc theory. Furthermore, the structure of cognitive abilities was invariant across sex, ethnic, and racial groups, and was replicated across age groups ranging from childhood to old age. Developmental evidence, in the form of differential growth curves has also discriminated among broad Gf–Gc abilities (e.g., Horn, 1985; Schaie, 1994). For example, some abilities appear to be ‘‘maintained’’ throughout ageing (e.g., comprehension-knowledge) while others are ‘‘vulnerable’’ to age-related declines (e.g., fluid reasoning). Clinically, decreases in vulnerable abilities have been related to traumatic brain injury and to chronic and acute alcohol

64

Waschbusch, Daleiden, and Drabman

use, whereas the maintained abilities may return to near premorbid functioning levels (Horn, 1989). Horn (1985; Horn & Noll, 1997) also cited a variety of evidence from genetic and neurocognitive studies supporting the distinctions between broad Gf–Gc abilities. For example, twin studies have indicated that the broad Gf and Gc abilities have different heritability estimates. Neurocognitive studies provide some indications that different brain regions and neurotransmitter systems may be differentially related to particular Gf–Gc abilities. Finally, Gf–Gc abilities have been related with a broader set of criterion variables that include ‘‘real life’’ outcomes and self-reported constructs. For example, various Gf–Gc abilities uniquely predict academic achievement variables (e.g., reading & math: McGrew, Keith, Flanagan, & Vanderwood, 1997, writing: McGrew & Knopik, 1993) and occupational outcomes (e.g., Dawis, 1994; Gardner, 1994). Finally, Ackerman & Heggestad’s (1997) recent meta-analysis of intellectual abilities, personality traits, and interests found evidence for several distinct ability-interest-personality clusters. Taken together, these data suggest that the components of Gf–Gc theory are reliably identifiable, display differential developmental trajectories, relate with distinct genetic, biological, and personality correlates, and predict several real life outcomes, thus supporting our choice of Gf–Gc theory as a basis for characterizing children’s cognitive abilities. There is an ongoing debate about whether the broad cognitive abilities described by Gf–Gc theory (or other theories of multiple intelligences) provide unique information over a single general ability (G) model. Positions relevant to this argument have ranged from the rejection of broad ability interpretation because it does not contribute information beyond that of general ability (e.g. Glutting, Youngstrom, Ward, Ward, & Hale, 1997; McDermott, Fantuzzo, & Glutting, 1990), through multistrata (Carroll, 1993) or hierarchical models of general and broad abilities (Gustafsson, 1989) to rejection of the notion of general ability as other than a mixture of distinct but related broad abilities (Horn, 1985). Despite this debate, researchers generally agree that measures can be selected and characterized relative to different broad ability classes, and that these various measures should display a high level of zero-order correlation. Accordingly, we examined parent reports of both broad Gf–Gc abilities and aggregated general ability (G). In the broad ability analyses we expected a high level of association (i.e., significant zero-order correlations) among parent reports, but we also investigated whether parent reports of each broad ability added unique information beyond general ability to the prediction of the criterion measures. In order to assess parental report, we developed behaviorally-specific items that we hypothesized to be indicators of these Gf–Gc factors. Our

Parent Report of Cognitive Abilities

65

hope in designing these items was to begin to understand what parents could and could not tell about their child’s cognitive abilities. We evaluated these items using a direct measure of children’s cognitive functioning, namely, the Woodcock-Johnson Psycho-Educational Battery-Revised (WJR; Woodcock, 1989). The WJ-R was selected as the primary measure for analyses because it was explicitly designed to measure Gf–Gc abilities. In addition, the perceptual organization and verbal comprehension index scores of the WISC-III were included because their subtests have been interpreted as indices of visual processing and comprehension-knowledge abilities (e.g., McGrew, 1997, Woodcock, 1990) and because they are widely used. The accuracy of parent ratings of their child’s cognition may differ as a function of their child’s sex. Potential sex differences in parent ratings are highlighted by data indicating that parent ratings of their child’s behavior yield higher problem rates, and sometimes different factor structures, in boys as compared to girls (Burns et al., 1997). Furthermore, boys and girls have demonstrated some consistent differences in their cognitive abilities, such that boys tend to have slightly higher visual-spatial perception skills and girls tend to have slightly higher verbal reasoning skills (Halpern, 1992). Differences in the accuracy of parent ratings of children’s cognitive abilities have yet to be examined even though finding such differences would have implications for the typical diagnostic interview. Therefore, the current study tested whether the accuracy of parent ratings differ as a function of their child’s sex. In sum, this paper explored whether parents are accurate reporters of their children’s cognitive abilities in a clinic-referred sample of youth. Gf–Gc theory was used as the theoretical basis of this investigation, and boys and girls were examined separately. Gf–Gc theory was operationalized using the WJ-R and the WISC-III to test children’s cognitive abilities, and behaviorally-based Likert ratings were used to test parent’s reporting abilities.

METHOD Participants Participants were 145 children consecutive referrals to an outpatient psychology clinic for assessment of cognitive abilities and behavioral functioning. Participants ranged in age from 5 to 12 (M ⫽ 8.04, SD ⫽ 2.06), with 89 boys (61.4% of sample) and 56 girls. Participants were predominantly Caucasian and middle or upper middle class. The majority of the partici-

66

Waschbusch, Daleiden, and Drabman

pants (81%) were in the non-clinical range (T-score ⬍65) on the total problem score (M ⫽ 59.57; SD ⫽ 9.97) of the child behavior checklist (Achenbach, 1991). A subset of the children (19%) was reported by their parents to be on medication. [Medication information was unavailable for 36 participants (25% of the sample).] The majority of the children on medication (85%) were taking stimulants. Medication was not discontinued for the testing session.

Materials Parent Ratings Parents rated 50 items designed to assess various aspects of their child’s behavior. Ratings were completed almost exclusively by mothers. Items were Likert scales ranging from 0 (‘‘not at all’’) to 3 (‘‘very much’’) and were generated from clinical experience, other ratings scales, and the behavioral tasks used in objective measures of cognitive assessment. A subset of these items was theorized to be representative of Gf–Gc factors. These items were combined into six rationally derived scales designed to index Gf–Gc abilities. Five of these six scales consist of four items that are designed to measure Gf–Gc specific ability factors (see Table I). The sixth scale is a summary score of all the items and is designed to measure general cognitive ability. Alpha reliabilities for these scales ranged from .60 to .89 and showed that each of the items for each of the scales should be retained (⬍.03 increment in alpha due to dropping single items). Scale scores were created by recoding items (so that higher scores indexed better cognitive abilities) and averaging within scales.

Woodcock-Johnson Psycho-Educational Battery—Revised The standard and supplemental batteries of the WJ-R tests of cognitive ability were designed to provide an operationalization of Gf–Gc theory. The WJ-R measures the six Gf–Gc factors used in this study (i.e., fluid reasoning, comprehension-knowledge, visual processing, auditory processing, short-term acquisition and retrieval, and long-term storage and retrieval) with standardized cluster scores based on multiple tests from the cognitive battery. These factors are combined into a general cognitive ability score, which was also used in this study. Reviews of the WJ-R have suggested good evidence for the fit of the WJ-R to Gf–Gc theory (Ysseldyke, 1990).

Parent Report of Cognitive Abilities

67

Table I. Parent Rated Items and Their Associated Scales Fluid Reasoning Scale (Alpha ⫽ .60) 1. Has difficulty grasping how the parts should fit together in a puzzle 2. Can not tell time with dial watch 3. Does s/he have trouble describing or remembering the way things s/he has seen look when they are not present? 4. If given two facts, does s/he have trouble integrating into a new concept? Comprehension-Knowledge Scale (Alpha ⫽ .66) 1. 2. 3. 4.

Cannot remember what s/he reads [is read to her/him for non-readers] Has difficulty following the conversations of others Does s/he have difficulty following verbal instructions? Does not seem to know how to talk to other children

1. 2. 3. 4.

Had trouble learning to cut with scissors Has problems drawing a straight line Artwork looks like it was done by someone much younger Seems very slow when using handwriting to copy material

1. 2. 3. 4.

Seems to have trouble hearing the difference between similar sounding words Seems to have trouble remembering which sounds go with which letters Seems to know fewer words than others his/her age Can read single words but gets confused when several words are together

1. 2. 3. 4.

When asked to do more than one thing at a time, tends to forget what to do next Has a difficult time remembering where things are Remembers things for a short time, then forgets them Has problems remembering the order in which events occurred

Visual Processing Scale (Alpha ⫽ .74)

Auditory Processing Scale (Alpha ⫽ .68)

Acquisition and Retrieval Scale (Alpha ⫽ .76)

General Cognitive Ability Scale (Alpha ⫽ .89) Average of all items

Wechsler Intelligence Scale for Children—Third Edition The WISC-III is a widely used, well-known assessment battery designed to measure the cognitive abilities of children ages 6 to 16. Factor analytic studies of the WISC-III (Wechsler, 1991) suggest a four-factor solution, two of which were used in this study: perceptual-organization and verbal comprehension. The perceptual-motor factor is thought to index visual-motor ability and the verbal-comprehension factor is thought to index comprehension-knowledge. Children’s general cognitive ability (i.e., full scale IQ score) from the WISC-III was also used. Participants younger than 6 years (n ⫽ 8) were administered the Wechsler Preschool and Primary Scale of Intelligence (Wechsler, 1967) in place of the WISC-III, therefore these cases were dropped from analysis of the WISC-III data.

68

Waschbusch, Daleiden, and Drabman

Procedures Participants in this study were assessed using a comprehensive battery that was administered by master- or doctoral-level psychologists (see Newcomb & Drabman, 1995, for details). The psychoeducational assessment battery was completed during two separate four-hour sessions that occurred on consecutive days. While the child participated in testing, a clinician interviewed the parents, which included asking parents to complete the items in Table I. Thus, the parent ratings and cognitive measures (i.e., WISC-III and WJ-R) were administered simultaneously by different individuals who did not have knowledge of the other’s results.

RESULTS First, Pearson correlations were used to examine the overlap between the parent rating scales. Separate correlations were computed for boys and girls. As shown in Table II, the parent rating scales were significantly associated for both boys and girls. Correlations for boys ranged from .24 to .83, and correlations for girls ranged from .09 to .83. Second, Pearson correlations were used to examine whether parents are valid reporters of their child’s general cognitive ability. As shown in rows 1 (for boys) and 7 (for girls) of Table III, the parent-rated General Cognitive Ability scale was significantly correlated with the children’s general cognitive ability as indicated by the WJR and the WISC-III. In order to examine sex differences, boys and girls correlations were compared using a Fisher r-to-z transformation. Results showed a nonsignificant difference for both with WJR (z ⫽ 0.15, ns) and for the WISC-III (z ⫽ 0.15, ns),

Table II. Pearson Correlations for Parent Rating Scale Scores with Other Parent Rating Scale Scores Computed Separately for Boys and Girls Scale 1. 2. 3. 4. 5. 6.

Fluid reasoning Comprehension-knowledge Visual processing Auditory processing Acquisition & retrieval General cognitive ability

FR

CK

VP

AP

AR

GCA

— .53*** .59*** .56*** .50*** .83***

.31* — .37*** .58*** .67*** .81***

.53** .32* — .24* .35** .69***

.34* .09 .19 — .46*** .72***

.40** .62*** .55*** .23 — .78***

.71*** .67*** .75*** .51*** .83*** —

Notes. Correlations for boys are in bottom left (n ⫽ 89) and correlations for girls are in top right (n ⫽ 56). FR ⫽ Fluid reasoning; CK ⫽ Comprehension-Knowledge; VP ⫽ Visual Processing; AP ⫽ Auditory Processing; AR ⫽ Acquisition and Retrieval; GCA ⫽ General Cognitive Ability. * ⫽ correlations significantly different from zero at p ⬍ .05; ** ⫽ p ⬍ .01; *** ⫽ p ⬍ .001.

General cognitive ability Fluid reasoning Comprehension-knowledge Visual processing Auditory processing Acquisition & retrieval

General cognitive ability Fluid reasoning Comprehension-knowledge Visual processing Auditory processing Acquisition & retrieval

.51*** .45** .31* .42** .38** .27*

.49*** .48*** .33** .29** .43*** .36** .43** .40** .18 .51*** .21 .24

.51*** .51*** .39*** .28* .46*** .38*** .38** .39** .21 .38** .13 .25

.34** .45*** .18 .19 .33** .15

Fluid reasoning WJR

.33* .42** .17 .23 .24 .15

.49*** .49*** .34** .22* .50*** .37*** .31* .33* .16 .28* .14 .20

.46*** .47*** .39*** .15 .50*** .33**

Comprehensionknowledge WJR WISC

.45*** .50*** .16 .48*** .26 .23

.33** .38*** .17 .38*** .24* .07

.42** .37** .17 .54*** .21 .21

.39*** .41*** .27* .26* .31** .30**

Visual processing WJR WISC

.26* .29* .14 .13 .35** .07

.30** .29** .28** .08 .29** .24*

Auditory processing WJR

.34* .40** .08 .26 .43** .08

.43*** .44*** .27* .18 .43*** .33**

.15 .18 ⫺.01 .16 .23 .005

.38*** .41*** .25* .31** .36*** .19

Acquisition & retrieval WJRa WJRb

Notes. Convergent validity correlations are boxed. a ⫽ Long term storage & retrieval; b ⫽ short term acquisition & retrieval. * ⫽ correlations significantly different from zero at p ⬍ .05; ** ⫽ p ⬍ .01; *** ⫽ p ⬍ .01.

7. 8. 9. 10. 11. 12.

Girls (n ⫽ 54 to 56)

1. 2. 3. 4. 5. 6.

Boys (n ⫽ 83 to 89)

Parent rating scales

General cognitive ability WJR WISC

Children’s cognitive test performance

Table III. Pearson Correlations Between Parent Ratings and Their Child’s Cognitive Test Performance Computed Separately for Boys and Girls

Parent Report of Cognitive Abilities 69

70

Waschbusch, Daleiden, and Drabman

suggesting that the validity of parent reports of boys and girls general cognitive abilities did not significantly differ. Examination of the magnitude of these correlations shows that parent reports explained about 20% to 25% of the variance in their child’s test performance. Third, Pearson correlations were also used to examine whether parents are valid reporters of broad cognitive abilities. As shown in rows 2 to 6 (for boys) and 8 to 14 (for girls) of Table III, parent ratings of boy’s broad cognitive abilities were all significantly correlated with the relevant criterion measure, with the exception of short term acquisition and recall on the WJR. In contrast, parent ratings of girls cognitive abilities were significant for fluid reasoning, visual processing, and auditory processing, but not for comprehension-knowledge or acquisition and retrieval. These zero-order correlations suggest that parents are valid reporters of their child’s broad cognitive abilities. However, these correlations do not distinguish the extent to which the observed relations are attributable to parent reports of general versus broad abilities. Fourth, multiple regression analyses were used to examine if parent ratings of broad cognitive abilities predicted children’s broad cognitive performance after controlling other parent rating scales. Regressions were computed separately for boys and girls. [We considered using sex as an independent variable rather than computing regressions separately for boys and girls. However, this strategy resulted in unacceptably high multicollinearity.] Parent rated scale scores of broad cognitive abilities (i.e., fluid reasoning, comprehension-knowledge, visual processing, auditory processing, and acquisition and retrieval) were entered as independent variables and children’s cognitive performance measures on these same broad cognitive abilities (i.e., WJ-R and WISC scores) were the dependent variables. With this strategy, the single degree of freedom statistics (i.e., the regression coefficients) provide an indication of whether parent-rated scales uniquely predicted each broad cognitive ability. Regressions are summarized in Table IV. Parent ratings of boys’ abilities accounted for significant variance in all of the cognitive test performance measures (see top of Table IV). Regression coefficients suggested that parent ratings of fluid reasoning and visual processing were uniquely associated with their child’s performance in these areas, but other ratings were not. However, the visual processing effect was only evident for the WJ-R not the WISC-III criterion. Unexpectedly, parent ratings of boys’ auditory processing seemed to function as a verbal comprehension index but not as an index of auditory processing. Thus, parent ratings of their boys included a significant component of general ability, but also included unique features for fluid reasoning and visual processing. Regression analyses showed a different pattern for girls (see bottom

Visual processing WJR WISC

Auditory processing WJR

Children’s cognitive test performance Comprehensionknowledge WJR WISC

Comprehensionknowledge WJR WISC

Visual processing WJR WISC

Auditory processing WJR

.26 .07 .16 .25 ⫺.22 (.24) 5, 83 5.12***

.27 .05 .16 .36* ⫺.23 (.29) 5, 50 4.09**

.09 ⫺.01 .17 .20 ⫺.16 (.09) 5, 50 0.98

Acquisition & retrieval WJRa WJRb

.34* ⫺.13 ⫺.10 .27* .16 (.27) 5, 83 6.04***

Acquisition & retrieval WJRa WJRb

Notes. Values in tables are standardized regression coefficients (betas). Convergent validity coefficients are boxed. a ⫽ Long term storage & retrieval. b ⫽ Short term acquisition & retrieval. * ⫽ p ⬍ .05; ** ⫽ p ⬍ .01; *** ⫽ p ⬍ .001.

Fluid reasoning .26 .37* .24 .31* .11 .21 Compreh.-know. .07 .10 .05 .03 .08 .16 Visual processing .23 .04 .14 .36* .56*** .008 Auditory process ⫺.01 .12 .02 .11 .09 .30* Acquis. & retriev. ⫺.03 ⫺.11 ⫺.002 ⫺.14 ⫺.22 ⫺.19 ......................................................................................................................................................................................... 2 (Equation R ) (.20) (.19) (.13) (.33) (.33) (.18) Equation df 5, 50 5, 48 5, 50 5, 50 5, 48 5, 50 Equation F 2.48* 2.40* 1.38 5.01*** 4.69** 2.17

Girls

Fluid reasoning WJR

Fluid reasoning .50** .33* .31* .25 .29 .22 Compreh.-know. ⫺.09 ⫺.07 .06 .001 ⫺.04 .12 Visual processing ⫺.08 ⫺.07 ⫺.15 .27* .06 ⫺.15 Auditory process. .16 .31* .30* .13 .09 .10 Acquis. & retriev. ⫺.08 .13 .02 ⫺.21 .10 .05 ......................................................................................................................................................................................... (Equation R2) (.23) (.32) (.31) (.22) (.18) (.13) Equation df 5, 83 5, 83 5, 77 5, 83 5, 77 5, 83 Equation F 5.01*** 7.92*** 6.77*** 4.61*** 3.32** 2.50*

Parent ratings boys

Fluid reasoning WJR

Table IV. Multiple Regression Analyses Examining Parent Ratings as Predictors of Their Children’s Cognitive Test Performance

Parent Report of Cognitive Abilities 71

72

Waschbusch, Daleiden, and Drabman

of Table IV). Parent ratings of girls’ abilities accounted for significant variance in test performance on one or more measures of fluid reasoning, visual processing, comprehension-knowledge, and acquisition and retrieval. Regression coefficients showed that parent-reported visual processing and auditory processing were uniquely related to their girl’s test performance on measures of these constructs. However, parent ratings of girls’ abilities tended to account for less total variance in the cognitive criteria. Thus, parent ratings for boys and girls seemed to perform quite differently from each other in relation to broad cognitive abilities.

DISCUSSION This study examined whether parents provided accurate information about their children’s cognitive abilities. Parental report was gathered using Likert ratings of behaviors that we hypothesized to be external expressions of children’s cognitive skills. Parent ratings were used to examine both general cognitive abilities (i.e., full scale IQ) and broad cognitive abilities (e.g., fluid reasoning, comprehension knowledge, etc.) and ratings were evaluated in terms of both general accuracy (i.e., simple correlations) and specificity (regression analyses). This study is one of the first to examine the validity of parents as informants of their children’s cognitive abilities. It is especially noteworthy that a rating scale was used to do so, as rating scales are an efficient and widely used method of gathering information from parents. Parent ratings of boys’ cognitive abilities showed good general accuracy as well as some specificity. Accuracy was evidenced in the simple correlations. Parent ratings of general cognitive ability were significantly correlated with children’s actual performance on the WJR and the WISC-III general ability scales. In fact, parent ratings explained approximately 25% of variance in these measures. Parent ratings of boys’ broad cognitive abilities were also significant in zero-order analyses, but the regression analyses suggest that these relations are accounted for by general ability. Specifically, regression analyses suggested that after taking into account other factors, parent report of fluid reasoning and visual processing demonstrated specificity, and the parent reported auditory processing functioned as a unique predictor of comprehension-knowledge. However, parent ratings of other abilities did not evidence specificity. These results suggest parents are valid reporters of their boys’ general cognitive ability and both valid and specific reporters of boys’ fluid reasoning and visual processing abilities. Parent ratings of girls’ cognitive abilities showed a less consistent pattern of results. The zero-order correlations showed that parents were accu-

Parent Report of Cognitive Abilities

73

rate reporters of girls’ general cognitive ability, explaining about 20% of the variance in girls’ actual cognitive performance. In addition, parents rating of girls’ broad cognitive abilities were significant in zero order correlations (i.e., were related to children’s actual performance), but only for fluid reasoning, visual processing, and auditory processing. Regression analyses suggested that after taking into account other factors, parent report of visual processing and auditory processing demonstrated specificity. That is, parent ratings of visual processing accounted for unique variance in the WJR and WISC visual processing measures. These results suggest parents are valid reporters of girls’ general cognitive ability and both valid and specific reporters of girls’ visual processing ability. Thus, parent ratings of both boys’ and girls’ cognitive abilities showed some evidence of specificity, although there were sex differences in which abilities demonstrated specificity. For both boys and girls, parent ratings of visual processing skills accounted for unique variance in visual processing performance, but only as measured by the WJ-R for boys. The sexes differed in that parent ratings of boys’ fluid reasoning skills were uniquely correlated with boys’ actual fluid reasoning abilities, but the same was not true for girls. However, parent ratings of girls’ auditory processing skills were uniquely correlated with girls’ actual auditory processing abilities, but the same was not true for boys. These results suggest that parents are able to identify their child’s cognitive strengths and weaknesses with some degree of specificity. The fact that parents were able to do so in this study supports the notion that there may be relatively direct behavioral correlates of cognitive abilities. When interpreting these results, it is important to bear in mind that both measure and method variances were represented in these findings. Significant correlations between parent and cognitive test measures required that both (a) the Likert scale adequately measured the construct (i.e., measure effect) and (b) parents were adequate reporters of that construct (i.e., method effect). Therefore, the findings that suggest a significant relationship between parent ratings and children’s test performance support the inferences that (a) the behaviors we chose are reasonable external indicators of children’s cognitive skills, and (b) that parents are accurate informants of their child’s abilities. In other words, the significant correlations for boys’ and girls’ general cognitive ability, fluid reasoning, visual processing, and auditory processing (see Table III) can be interpreted as showing that the behaviors we used as indicators of these skills (see Table I) are reasonably valid. In addition, the results suggest that parents can report on these behaviors with moderate validity. On the other hand, not all scales received empirical support. That is, neither parent ratings of comprehension-knowledge nor parent ratings of

74

Waschbusch, Daleiden, and Drabman

acquisition and recall (i.e., memory) skills accounted for unique variance in any criterion for either sex. There are at least two alternative hypotheses for these results. First, these scales may be poor measures of their respective constructs regardless of whether parents are reasonable informants of these constructs. Second, regardless of whether these scales are adequate measures of the constructs, parents may be invalid reporters of their children’s comprehension-knowledge and memory abilities. If true, this second hypothesis would have important clinical implications. For example, the most recent revision of the DSM (from DSM-III-R to DSM-IV) added a symptom to ADHD that specifically assesses memory problems (‘‘is often forgetful in daily activities’’). If our results are correct, and it is parents who are not valid reporters of their child’s memory ability (rather than our scale being an invalid tool), then symptoms indexing memory—such as this ADHD symptom – should not be assessed using parent reports. Future studies will be necessary to evaluate these competing explanations. While the present analysis provides useful information for interpreting parental reports of children’s cognitive abilities, we are not advocating substitution of parent ratings for performance-based screening or standardized cognitive testing. Performance tests are the ‘‘gold standard’’ for assessing a child’s general or broad cognitive abilities and carefully selected subtest batteries can provide quite inexpensive and accurate estimates of cognitive ability. However, for the reasons noted in the introduction, performance tests are not always administered prior to drawing clinical inferences about children’s abilities. Instead, parent reports of children’s cognitive abilities are often elicited and used in both clinical and educational settings. Our results suggest that parents can provide a reasonably accurate indication of their child’s general cognitive ability as compared to other children, but the accuracy of parental report for particular aspects of the child’s cognitive abilities is less certain. Specifically, our results suggest that educators/clinicians can be most confident of the accuracy and specificity of parent reports of: (1) boys ability to reason in novel situations (i.e., fluid reasoning skills); (2) girls ability to analyze sound and language; and (3) both boys and girls and boys ability to perceive and think with visual patterns (i.e., visual processing skills); and (4) boys and girls overall cognitive ability level. However, when educators and clinicians ask parents about other skills, our results suggest that they should be more cautious about how they interpret and use what parents report. There are a number of limitations to this study that should be noted. First, the sample used was one of convenience, comprised of children and their parents who presented to our clinic for a variety of psychological or behavioral problems. It is unclear how this fact influences the results, but these children were representative of the clientele routinely presenting to

Parent Report of Cognitive Abilities

75

a mental health clinic. Second, boys were over-represented in our sample because more boys than girls presented for treatment. This pattern is commonly reported and likely reflects a higher prevalence of adjustment problems among boys than girls (at least in childhood). Third, the parent rating scale represented an early attempt to collect parent reports of children’s broad cognitive abilities and included few items per scale. As a result, the internal consistencies (i.e., alpha reliabilities) of some scales were less than .70. Even so, the fact that these preliminary data suggest that parents can validly report on child cognitive abilities highlights the promise of future research using carefully constructed instruments with more items per scale. Taken together, these limitations and the novelty of this study emphasize the need to replicate results before drawing firm conclusions. Replication is especially important for the results involving fluid reasoning and auditory processing as the present study used only one performance-based criterion measure to assess each of these scales. Research examining the generalization of these results using a non-clinical sample would also be beneficial, as would expanding the items included in the scales. In summary, this study examined whether parents can accurately report their child’s cognitive abilities. Results suggested that parents can accurately report boys’ and girls’ general cognitive ability, fluid reasoning, visual processing, and auditory processing. Finally, results suggested that parents were specific reporters of boys’ fluid reasoning as well as girls’ visualprocessing abilities. Thus, the performance of parent reports differed depending on their child’s sex. Although these conclusions must be considered preliminary until replicated in other samples, the results provide important information to clinicians and educators about use of parental information to evaluate children’s cognitive functioning.

ACKNOWLEDGMENT This project was supported in part by a Faculty Development Grant from the University of Tulsa and by a Faculty Development Grant from Dalhousie University.

REFERENCES Achenbach, T. M. (1991). Manual for the Child Behavior Checklist (ages 4–18) and 1991 Profile. Burlington, VT: University of Vermont Department of Psychiatry. Achenbach, T. M., McConaughy, S. H., & Howell, C. T. (1987). Child/adolescent behavioral and emotional problems: Implications of cross-informant correlations for situational specificity. Psychological Bulletin, 101, 213–232.

76

Waschbusch, Daleiden, and Drabman

Ackerman, P. L., & Heggestad, E. D. (1997). Intelligence, personality, and interests: Evidence for overlapping traits. Psychological Bulletin, 121, 219–245. Burns, G. L., Walsh, J. A., Patterson, D. R., Holte, C. S., Sommers-Flanagan, R., & Parker, C. M. (1997). Internal validity of the disruptive behavior disorder symptoms: Implications from parent ratings for a dimensional approach to symptom validity. Journal of Abnormal Child Psychology, 25, 307–319. Carroll, J. B. (1993). Human cognitive abilities: A survey of factor analytic studies. New York: Cambridge University Press. Dawis, R. V. (1994). Occupations. In R. J. Sternberg (Ed.), Encyclopedia of human intelligence (pp. 781–785). New York: Macmillan. Fabrega, H. J., Ulrich, R., & Loeber, R. (1996). Adolescent psychopathology as a function of informant and risk status. Journal of Nervous and Mental Disease, 184, 27–34. Gardner, H. (1994). Multiple intelligences theory. In R. J. Sternberg (Ed.), Encyclopedia of human intelligence (pp. 740–742). New York: Macmillan. Glutting, J. J., Youngstrom, E. A., Ward, T., Ward, S., & Hale, R. L. (1997). Incremental validity of WISC-III factor scores in predicting achievement: What do they tell us? Psychological Assessment, 9, 295–301. Gustafsson, J. E. (1989). Broad and narrow abilities in research on learning and instruction. In R. Kanfer, P. Ackerman, & R. Cudeck (Eds.), Abilities, motivation, and methodology: The Minnesota symposium on learning and individual differences (pp. 203–237). Hillsdale, NJ: Erlbaum. Halpern, D. F. (1992). Sex differences in cognitive ability. (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum. Hart, E. L., Lahey, B. B., Loeber, R., & Hanson, K. S. (1994). Criterion validity of informants in the diagnosis of disruptive behavior disorders in children: A preliminary study. Journal of Consulting and Clinical Psychology, 62, 410–414. Horn, J. L. (1985). Remodeling old models of intelligence. In B. B. Wolman (Ed.), Handbook of intelligence: Theories, measures, and applications (pp. 267–300). New York: Wiley. Horn, J. L. (1989). Cognitive diversity: A framework of learning. In P. L. Ackerman, R. J. Sternberg, & R. Glaser (Eds.), Learning and individual differences: Advances in theory and research (pp. 61–116). New York: Freeman. Horn, J. L. (1994). Theory of fluid and crystallized intelligence. In R. J. Sternberg (Ed.), Encyclopedia of human intelligence (pp. 443–451). New York: Macmillian. Horn, J. L., & Noll, J. (1997). Human cognitive capabilities: Gf-Gc theory. In D. P. Flanagan, J. L. Genshaft, & P. L. Harrison (Eds.), Contemporary intellectual assessment: Theories, tests and issues (pp. 53–91). New York: Guilford. Loeber, R., Green, S. M., Lahey, B. B., & Stouthamer-Loeber, M. (1990). Optimal informants on childhood disruptive behaviors. Development and Psychopathology, 1, 317–337. Loeber, R., Green, S. M., Lahey, B. B., & Stouthamer-Loeber, M. (1991). Differences and similarities between children, mothers, and teachers as informants on disruptive child behavior. Journal of Abnormal Child Psychology, 19, 75–95. McDermott, P. A., Fantuzzo, J. W., & Glutting, J. J. (1990). Just say no to subtest analysis: A critique on Wechsler theory and practice. Journal of Psychoeducational Assessment, 8, 290–302. McGrew, K. S. (1997). Analysis of the major intelligence batteries according to a proposed comprehensive Gf-Gc framework. In D. P. Flanagan, J. L. Genshaft, & P. L. Harrison (Eds.), Contemporary intellectual assessment: Theories, tests, and issues (pp. 151–174). New York: Guilford. McGrew, K. S., Keith, T. Z., Flanagan, D. P., & Vanderwood, M. (1997). Beyond g: The impact of Gf-Gc specific cognitive abilities research on the future use and interpretation of intelligence tests in the schools. School Psychology Review, 26, 189–210. Newcomb, K. P., & Drabman, R. S. (1995). Child behavioral assessment in the psychiatric setting. In R. T. Ammerman & M. Hersen (Eds.), Handbook of child behavior therapy in the psychiatric setting (pp. 3–25). New York: John Wiley & Sons. Sattler, J. M. (1992). Assessment of children. (3rd revised ed.). San Diego: Jerome M Sattler Publisher.

Parent Report of Cognitive Abilities

77

Schaie, K. W. (1994). The course of adult development. American Psychologist, 49, 304–314. Wechsler, D. (1967). Wechsler Preschool and Primary Scale of Intelligence. New York: Psychological Corporation. Wechsler, D. (1991). Wechsler Intelligence Scale for Children—Third Edition. New York: Psychological Corporation. Woodcock, R. W. (1989). Woodcock-Johnson Psycho-Educational Battery—Revised. Allen, TX: DLM Teaching Resources. Ysseldyke, J. E. (1990). Goodness of fit of the Woodcock-Johnson Psycho-Educational Battery—Revised to the Horn-Cattell Gf-Gc theory. Journal of Psychoeducational Assessment, 8, 268–275.

Suggest Documents