Reliability and Validity the Brief Problem Monitor, an Abbreviated Form of the Child Behavior Checklist

Abstract

Aim

The parent form of the 113 item Child Behavior Checklist (CBCL) is widely utilized by child psychiatrists and psychologists. This report examines the reliability and validity of a recently developed abbreviated version of the CBCL, the Brief Problem Monitor (BPM).

Methods

Caregivers (N=567) completed the CBCL online and the 19 BPM items were examined separately.

Results

Internal consistency of the BPM was high (Cronbach’s alpha=0.91) and satisfactory for the Internalizing (0.78), Externalizing (0.86), and Attention (0.87) scales. High correlations between the CBCL and BPM were identified for the total score (r=0.95) as well as the Internalizing (0.86), Externalizing (0.93), and Attention (0.97) scales. The BPM and scales were sensitive and identified significantly higher behavioral and emotional problems among children whose caregiver reported a psychiatric diagnosis of Attention Deficit Hyperactivity Disorder, bipolar, depression, anxiety, developmental disabilities, or Autism Spectrum Disorders relative to a comparison group that had not been diagnosed with these disorders. BPM ratings also differed by the socioeconomic status and education of the caregiver. Mothers with higher annual incomes rated their children as having 38.8% fewer total problems (Cohen’s d=0.62) as well as 42.8% lower Internalizing (d=0.53), 44.1% less Externalizing (d=0.62), and 30.9% decreased Attention (d=0.39). A similar pattern was evident for maternal education (d=0.30 to 0.65).

Conclusion

Overall, these findings provide strong psychometric support for the BPM although the differences based on the characteristics of the parent indicates that additional information from other sources (e.g., teachers) should be obtained to complement parental reports.

Introduction

The Child Behavior Checklist (CBCL) is a caregiver completed questionnaire of child behavioral and emotional problems which is standardized, objective, and widely utilized by child psychiatrists, pediatricians, developmental psychologists, and other mental health professionals for clinical and research purposes.¹ The CBCL has been revised since its original introduction by Thomas Achenbach, Ph.D. and colleagues and there are various forms of the instrument available depending on the information source (parent, teacher, and self-report), language of the respondent, and child age (preschool or school age). The CBCL/6-18 parent-report has high test-retest reliability, internal consistency, criterion validity and shows good agreement between maternal and paternal ratings.^2–6 The influence of differences in the socioeconomic environment to CBCL ratings is a matter of some contention.^3,7,8 Various investigators are in general agreement that children living in more economically disadvantaged households are rated as having more problems but the magnitude of this effect has been reported as either modest⁹ or, more commonly, as sizable.^10–12

One potential concern with the CBCL/6-18 is its length. There are 113 problem items which take approximately ten minutes to complete and the optional competence items require another five to ten minutes. The development of an abbreviated version of the CBCL offers several advantages. First, a shorter form allows clinicians to more easily determine the potential presence of behavioral and emotional problems. This could involve repeated measurement of a single child to assess the utility of a pharmacological or psychoeducational intervention. Second, this instrument could provide non-specialists with a screening tool to identify children and adolescents for whom follow-up with a specialist is appropriate (e.g. for a neurodevelopmental disorder, schizophrenia, bipolar, etc.). Third, a shortened version would be useful for investigators interested in learning about individual difference in emotion and behavioral functioning but who view the full-length form as prohibitively long within the context of multiple data collection demands (e.g. an epidemiological neurogenetics study).

The Brief Problem Monitor (BPM) Parent form was introduced in 2011 by Achenbach and coworkers and can be completed in only two-minutes as it consists of only 19 items (Appendix A) from the CBCL which form scales for Internalizing, Externalizing, and Attention.¹³ The BPM is a modification and extension of the 12-item Brief Problem Checklist (BPC).¹⁴ The six items that form the Internalizing and Externalizing scales of the BPC were identified via application of item response theory and factor analysis to a large archival CBCL dataset. Three-month test-retest reliabilities of the BPC when the instrument was administered in an interview format were satisfactory (r = .73 to .82). An advantage of the BPM over the BPC is that the former contains additional items to assess attention. The prevalence of Attention Deficit Hyperactivity Disorder (ADHD, 5%)¹⁵ is sufficiently high to warrant the addition of the six additional items to the BPC for the BPM (Table 2 of citation¹⁴ contains a complete list of BPC items). The BPM¹³ also includes an additional Externalizing item on disobedience. Therefore, the objective of the present report was to build and expand upon the earlier work and examine several psychometric properties of the BPM including internal consistency and validity, similarities with the CBCL, and also determine if there are any individual differences associated with demographic variables (e.g. education of the respondent).

Materials and Methods

Procedures

Research Electronic Data Capture (REDCap), version 1.3.9, a web-based program for obtaining online databases which has multiple safeguards for protecting personal information¹⁶, administered the anonymous survey. Flyers were posted throughout the Oregon Health and Science University (OHSU) and Doernbecher Children’s hospitals, the Portland, Oregon metro area, and western Oregon and western Washington (e.g. coffee shops, laundromats, grocery stores). Links to the investigation were placed on the community and volunteer sections of Craigslist (craigslist.org) in the US and Canada for one year. The Institutional Review Boards of OHSU and Northern Arizona University approved all procedures which were in accordance of the Declaration of Helsinki.

Measures

After completing an online consent to complete this investigation, caregivers began the survey which took approximately 20 minutes to finish. Items on the first half (Appendix B) included maternal and child demographics (age, sex, ethnicity, education and household income). “Has your child ever been diagnosed with any of the following?” was listed with 28 options including common psychiatric and medical conditions (e.g. ADHD, ASD, Attachment Disorder, bipolar, Developmental Disorder, Diabetes, Schizophrenia). Data about paternal characteristics were not obtained in this initial study due to a desire to minimize the survey completion time for each participant. Information from items about maternal drug use and child outcomes is presented elsewhere¹⁷. The last half of the survey consisted of the CBCL. Caregivers were instructed to rate over the past half-year whether an item (e.g. inattentive) was not, somewhat, or very true about their child with scores of 0, 1, or 2 points.

Data Analysis

All analyses were completed in Systat, v 13.0 (Chicago, IL). Exclusionary criteria included omitting items about the child age, sex; child age (> 18 or <6), or not completing more than one BPM item. Note that although REDCap has the capability to require responses, the only required item was the consent in order to keep the procedures as similar as possible to traditional (i.e. paper and pencil) administration. The 19 items that make up the BPM (Appendix A) were isolated and scale scores calculated. Internal consistency of the BPM and each scale was obtained using Cronbach’s alpha. A common standard for internal consistency is that Cronbach’s alpha values should be greater than 0.7¹⁸. Pearson product moment correlations (r) and the proportion of variance accounted for (R²) were calculated between the BPM and CBCL as examining the relationship between the full-length and abbreviated instrument is common step in test development^19–22. Respondents (N = 28) who selected the “unknown/prefer not to disclose” option for the demographic item on annual income were removed from subsequent income analyses. As the demographic items focused on the birth mother (e.g. What is the highest level of schooling that the biological mother completed?) analyses of these variables and their relationships to the BPM were restricted to biological mother caregivers. Non-parametric (chi-square) analyses were conducted with BPM total and scale scores that were categorized as high (T₅₀ ≥ 65). A non-clinical comparison group was defined as children who had not been diagnosed with ADHD, depression, bipolar, anxiety including Post-Traumatic Stress Disorder (PTSD), an Autism Spectrum Disorder (ASD), a developmental disability, Fetal Alcohol Syndrome, and cerebral palsy. Construct validity was evaluated by examining with a t-test if the generally expected BPM elevations were observed (e.g. Do children diagnosed with ADHD, relative to the comparison group, show significant increases in Attention Problems?). As online research studies in child psychiatry are only gradually^17,23 becoming more widely utilized, the prevalence of comorbidities was examined to determine if this dataset showed the expected patterns as has been identified with other methodologies¹⁵. An alpha of < .05 was considered significant although statistics that met more conservative thresholds were noted. Group differences were also expressed in terms of Cohen’s d with 0.2 considered small, 0.5 medium, and 0.8 large effect sizes.

Results

The vast majority of caregivers completing the study (N = 567) were the biological mother (82.3%) followed by adoptive/foster parents (10.9%), the biological father (4.6%) or other family members (2.1%). Many respondents were from the western U.S. (Oregon = 36.4%, Washington = 13.0%, California = 11.8%, Arizona = 3.8%, Idaho = 2.5%) or other regions (Canada = 4.5%, Texas = 3.6%, New York = 3.4%, Wisconsin = 2.3%). Among biological mother participants, two-fifths had been diagnosed with depression (41.0%), over one-quarter had an anxiety disorder including PTSD (26.6%) followed by Bipolar (7.5%), Diabetes (6.4%), ADHD (4.7%), and epilepsy (2.1%). The highest level of education completed by the mother was high school or a general equivalency diploma for approximately one-quarter (26.3%) of the sample with three-fifths attending or completing college (58.4%), and the remaining participants (15.3%) having a graduate or professional education. Annual household income was above $50,000 for almost half (45.0%) and below $20,000 for one-sixth (17.7%) of caregivers.

Half the children were female (50.1%) with an average age of 11.5 ± 0.2 (6 to 9: 41.3%, 10 to 13: 29.5%, 14 to 18: 29.3%); primarily white (75.2%), Hispanic (7.6%), African-American (2.8%), American Indian (3.4%), or other (11.0%) ethnicity. Children were attending public school (83.2%), private school (5.7%), an alternative education program (4.4%), or were home-schooled (3.9%). An appreciable minority of children had been diagnosed with ADHD (17.4%) followed by anxiety (12.7%), depression (9.5%), an Attachment Disorder (5.0%), Post-Traumatic Stress Disorder (4.4%), a Developmental Disability (4.4%), Autism Spectrum Disorder (4.1%), Bipolar Disorder (3.2%), Fetal Alcohol Syndrome (1.6%), Tourette’s Syndrome (0.9%), epilepsy (0.7%), or Cerebral Palsy (0.5%).

The internal consistency of the BPM was quite satisfactory with a Cronbach’s α of 0.91 for all items, 0.79 for Internalizing, 0.86 for Externalizing, and 0.87 for the Attention Problems scales. The consistency of biological parents was compared to adoptive/foster parents and found to be quite similar (Supplemental Table 1). There were also no appreciable differences based on the sex or age of the child (Supplemental Table 2).

The correspondence between the BPM and the CBCL was excellent for the total scores (R²(510) = .90), Internalizing (R²(565) = 0.74), Externalizing (R²(565) = 0.86), and Attention Problems (R²(565) = 0.94) scales (Figure 1).

Figure 1.

Scatterplots with linear regression between the parent forms of the full-length Child Behavior Checklist and the abbreviated Brief Problem Monitor items for the Total (r(510) = 0.95, p < .0005), Internalizing (r(547) = 0.87, p < .0005), Externalizing (r(547) = 0.93, p < .0005), and Attention Problems (r(555) = 0.97, p < .0005) scales.

Table 1 shows that children and adolescents with a variety of psychiatric diagnoses including ADHD, mood disorders, Attachment Disorder, or an ASD had a total BPM that was at least doubled relative to a non-diagnosed comparison group. Caregivers of ADHD children reported a three-fold increase in Attention Problems. Internalizing Problems were similarly more common among children with Depression (3.6 fold) or an Anxiety Disorder (3.2 fold). Children with Bipolar had substantial increases in Externalizing (3.6 fold), Internalizing (3.4 fold), and Attention (3.2 fold) problems.

Table 1.

Brief Problem Monitor total and scale scores among children diagnosed with Attention Deficit Hyperactivity Disorder, Depression, an Anxiety Disorder (including Post-Traumatic Stress Disorder N = 25), an Attachment Disorder (Attat Dis), a Developmental Disorder (Dev Dis), Autism Spectrum Disorder (ASD, including autism N = 14), and a Speech Delay.

	Comparison	ADHD	Depression	Anxiety	Bipolar	Attat Dis	Dev Dis	ASD	Speech Delay
	N = 384	N = 76	N = 54	N = 82	N = 18	N = 28	N = 25	N = 23	N = 44
Total	6.5 (0.3)	16.7 (0.7)CC	17.6 (1.0)CC	16.6 (0.9)CC	22.1 (1.4)CC	17.8 (1.5)CC	17.1 (1.6)CC	14.5 (1.6)CC	13.5 (1.2)CC
Internalizing	1.5 (0.1)	2.9 (0.3)CC	5.4 (0.4)CC	4.8 (0.3)CC	5.1 (0.7)CC	3.9 (0.5)CC	3.8 (0.6)CC	3.5 (0.6)C	2.6 (0.4)C
Externalizing	2.5 (0.1)	6.0 (0.4)CC	6.2 (0.5)CC	5.7 (0.5)CC	9.2 (0.9)CC	7.0 (0.8)CC	6.1 (0.7)CC	4.6 (0.7)C	4.5 (0.5)C
Attention	2.5 (0.1)	7.8 (0.3)CC	6.1 (0.5)CC	6.1 (0.4)CC	7.9 (0.6)CC	6.8 (0.6)CC	7.1 (0.8)CC	6.4 (0.7)CC	6.4 (0.6)CC

^Cp < .05 or

^CCp ≤ .0005 t-test versus the comparison group

Further analyses revealed the anticipated high comorbidities with over one-quarter of the ADHD group also diagnosed with Depression (28.9%) and almost half having an Anxiety Disorder (47.4%). The majority of the Bipolar children also had ADHD (61.1%) or an Anxiety Disorder (55.6%). Two-thirds of children with Depression (66.7%) also had an Anxiety Disorder. Three-quarters with an Attachment Disorder had an Anxiety Disorder and over half also had Depression (53.6%) or ADHD (53.6%). Less than one-third of offspring with a Speech Delay also had a Developmental Disability (29.5%), ADHD (27.3%) or an Anxiety Disorder (22.7%) and approximately only one-tenth had an ASD (13.6%) or Depression (9.1%).

Finally, analyses were conducted to determine if demographic variables contributed to scores on the BPM. Among biological mothers, those with a college education reported 20.5% fewer Externalizing Problems in their progeny relative to those with a high-school education (t(391) = 2.28, p < .05, d = .24) but only 15.5% less Total Problems (t(204.4) = 1.85, p = .066). Mothers with a professional or graduate education also related 43.2% less Total Problems (t(189.2) = 4.56, p ≤ .0005, d = 0.64), a 30.0% reduction in Internalizing Problems (t(191) = 1.98, p < .05, d = 0.30), 48.7% less Externalizing Problems (t(188.7) = 4.60, p < .0005, d = 0.65), and 45.5% fewer Attention Problems (t(188.6) = 2.56, p ≤ .0005, d = 0.60) than women with a high-school education. Biological mothers with professional/graduate education, as compared with respondents with a college education, also rated their children as having significantly fewer Total problems (t(143.3) = 3.76, p ≤ .0005, d = 0.46), Externalizing Problems (t(147.3) = 3.37, p ≤ .001, d = 0.41), and Attention Problems (t(151.0) = 3.96, p ≤ .0005, d = 0.47, Figure 2A). Similar to the pattern of findings for education, biological mothers with higher annual house-hold annual income (≥$20K to 49.9 K/year) rated their offspring as having 22.5% fewer Total (t(242) = 2.67, p < .01, d = .40), 30.0% less Internalizing (t(242) = 2.63, p < .01, d = 0.35), and a 24.0% decrease in Externalizing (t(242) = 2.46, p < .05, d = .33) problems relative to mothers with lower incomes (< $20K/year). Children with biological mothers with even higher incomes (> $50 K/year) had a 38.8% decrease on the BPM total (t(274) = 4.84, p ≤ .0005, d = 0.62), as well as a 42.8% reduction for Internalizing (t(274) = 4.23, p < .0005, d = 0.53), a 44.1% decrease for Externalizing (t(274) = 4.83, p ≤ .0005, d = 0.62), and a 30.9% less for Attention (t(274) = 3.03, p < .005, d = 0.39, Figure 2B). The contribution of education and family income to BPM ratings was also identified among respondents with (Supplemental Figure 1), and without (Supplemental Figure 2), a major mental illness. Supplemental Table 3 shows a very similar pattern for education and income when the data was expressed as the percent of children that scored above the clinical cutoff (T₅₀ ≥ 65). No appreciable BPM differences were noted based on the ethnicity of the respondent (data not shown) or the child (Supplemental Table 4).

Figure 2.

A) Brief Problem Monitor (BPM) total and scale scores by education of the biological mother expressed as a percentage of the high-school group (Total = 9.9, SEM = 0.7; Internalizing = 2.3, SEM = 0.2; Externalizing = 3.8, SEM = 0.3; Attention = 3.8, SEM = 0.3). ^ap < .05 versus high-school, ^bp < .05 versus college. B) BPM by current annual income (in thousands) expressed as a percentage of the < 20 K/year group (Total = 11.2, SEM = 0.8; Internalizing = 2.9, SEM = 0.3; Externalizing = 4.3, SEM = 0.3; Attention = 4.1, SEM = 0.4). ^cp < .05 t-test versus <20 K/year, ^dp < .05 t-test versus 20 to 49 K/year.

Discussion

The CBCL has been extensively utilized as a clinical and research instrument for a wide-variety of psychiatric, neurological, and other medical conditions including brain damage, Oppositional disorder, cancer, cystic fibrosis, Prader-Willi syndrome, lead toxicity, HIV, and epilepsy¹. Instruments like the CBCL are important because of the increased recognition of the many disorders where symptom expression may occur prior to adulthood. This includes not only ADHD, ASD, Conduct Disorder, and learning disabilities but also some anxiety disorders, Bipolar I Disorder, and Paranoid Personality Disorder^15,24. One limitation is that the parent form of the CBCL consists of 113 items, including one item with several parts, and may require at least ten minutes to complete. This report shows that the BPM (Appendix A), despite being only one-sixth as long as the original, has excellent psychometric properties including good internal consistency and high correlations with the full-length CBCL including the ability to differentiate children with various psychiatric diagnoses. These findings indicate that the short-form’s composite and scales are measuring substantively the same constructs as the CBCL. The BPM, like the CBCL,¹ will likely be of substantial use in applied and experimental environments for use with children and adolescents with ADHD, anxiety, depression, as well as many other neurobehavioral conditions.

It is rather interesting that the BPM total as well as the Internalizing, Externalizing, and Attention scales were significantly elevated among children and adolescents with a wide array of conditions including Bipolar, an Attachment Disorder, Developmental Disabilities or an Autism Spectrum Disorder, relative to a comparison group without these disorders. This pattern of results suggests that the BPM may have very good sensitivity. The pronounced increase in Externalizing among the Bipolar children, in Attention problems in ADHD children, and in Internalizing among the anxious and depressed groups is consistent with predictions. Further, prior research with children with Language Development Disorders show broad CBCL increases²⁵ which is congruent with the present findings with a Speech Delay. However, additional study with children with fewer comorbid conditions will be informative in determining the specificity of the BPM as specificity was limited in this investigation.

Although the finding that maternal characteristics contribute to BPM ratings was not unanticipated²⁶, we were surprised by the magnitude of this effect. The effect size was at least moderate (d ≥ .60) for maternal education and annual income for the BPM total and low to moderate (d = 0.30 to 0.64) for each scale. As has been noted by others⁷, one might reasonably infer from reading the original CBCL manual that the contribution of SES to parental ratings is small or even trivial. Achenbach and Rescorla⁶ very briefly note that 23 CBCL items do show higher scores for lower SES children but also that the proportion of variance uniquely accounted for by SES for each item was rather small. In contrast, investigators that have used the CBCL with a disadvantaged urban sample discovered that over one-quarter of low-SES children scored in the clinical/borderline range.¹¹ Similarly, a difference of 0.7 standard deviations in the CBCL total score between offspring of Dutch parents with high and low education (d = 0.61) was observed with only a slightly smaller effect size (d = 0.54) for parental occupational status.¹⁰

One caveat with the CBCL, which also likely applies to the BPM, is that mental health of the respondent may be associated with elevated parental ratings of their children. Mothers who were anxious and depressed were four times more likely to have children with CBCL problems.²⁷ The agreement among different informants about the child’s behavior in different circumstances (e.g. parent versus child versus teacher) is often only moderate at best.^5,18 We would like to re-emphasize that the BPM is only one source of information which, given the sizable group differences depicted in Figure 2, should be complemented, whenever possible, with further data from teachers and the child²⁸.

An earlier effort to develop a short-form of the CBCL, the BPC¹⁴, forms a crucial foundation for the present study but a couple methodological differences should also be highlighted. The BPC consists of the six Internalizing and six Externalizing items that are also contained in the BPM but does not contain the Attention items. The BPM, but not the BPC, is currently supported by the Achenbach System of Empirically Based Assessment, because interventions often target problems with inattention and overactivity¹³ and additional focus on that domain was warranted. Therefore, there is a much greater likelihood that the BPM will become the standard abbreviated form of the CBCL. The mode of administration also differed between investigations with Chorpita and colleagues using face to face interviews¹⁴ while this study employed online delivery. Electronic survey administration is potentially advantageous over more traditional (i.e. paper and pencil) methods in that calculation of scaled scores can be automated and participants that might be difficult to interview (e.g. families in rural areas) can readily be included. Alternatively, a possible disadvantage is that respondents in their home-environments might be unable to devote their undivided attention to filling out the instrument or be less likely, once started to complete an online survey. For a much more expansive discussion of the pros and cons of online data-collection with convenience samples see²⁹. As the Cronbach’s alphas for the BPM total and scales were above generally accepted guidelines¹⁹ and approximately equivalent to the Brief Problem Checklist (.05 higher for Externalizing but .05 lower for Internalizing),¹⁴ these results imply that caregivers can provide data that is at least as internally consistent as that obtained through other methods. These findings attest to the many advantages of information gathered electronically, particularly for material of a sensitive nature, relative to other methodologies.³⁰

Three limitations and future directions should also be noted. First, although the sample was moderately large and contained a substantial number of both respondents and children with diverse psychiatric conditions, the participants were primarily mothers who were also uncompensated volunteers (i.e. they were self-selected) from the western U.S.. The goal of this report was to include respondents that would similar to those that will be completing the BPM in the future. Relative to the U.S. census, the present sample was ethnically diverse (24.9% non-white versus 22.1% from the census) but also included a disproportionate number of respondents from a low socioeconomic background. As the CBCL is currently available in over 90 languages, we anticipate that many additional studies will be forthcoming which evaluate alternative forms of the BPM with samples from many other countries. Second, this report relied on maternal reporting of whether children had been diagnosed with various psychiatric disorders. This raises the possibility of under/over reporting or misdiagnosis. Future studies could at least partially address this concern by making greater use of other information sources (fathers, grandparents, or teachers) or by examining medical records to corroborate the parental reports. Third, while this is the first study to examine the psychometric properties of the BPM, further research is warranted. One potential concern with evaluating the criterion validity of the BPM with items extracted from the CBCL is overlapping measurement error which could inflate the correlation. Unfortunately, there is not a single formula which is uniformly accepted to correct for this but, when formulas are applied³¹, the magnitude of reduction is relatively modest (.03). Nevertheless, the present findings in conjunction with prior data using instruments with scales that overlap with the BPM but with the CBCL administered separately¹⁴ attest to the strong criterion validity of this measure. Future objectives include evaluating the test-retest reliability of the BPM as well as comparing results with those obtained with other behavioral screening questionnaires.³²

Overall, these findings provide strong psychometric support for the parent form of the BPM. The BPM offers a reliable and valid option when an abbreviated from of the CBCL is appropriate.

Supplementary Material

Supp Figure 01-02 & Table 01-04

Supplemental Figure 1. A) Brief Problem Monitor (BPM) total and scale scores by education among biological mother respondents not diagnosed major mental illness (N = 265) expressed as a percentage of the high-school group (Total = 7.5, SEM = 0.9; Internalizing = 1.8, SEM = 0.3; Externalizing = 2.7, SEM = 0.4; Attention = 3.1, SEM = 0.4). ^ap < .05 versus high-school, ^bp < .05 versus college. B) BPM by current annual income (in thousands) expressed as a percentage of the < 20 K/year group (Total = 10.7, SEM = 1.4; Internalizing = 2.5, SEM = 0.5; Externalizing = 3.9, SEM = 0.5; Attention = 4.3, SEM = 0.6). ^cp < .05 t-test versus <20 K/year, ^dp < .05 t-test versus 20 to 49 K/year.

Supplemental Figure 2. A) Brief Problem Monitor (BPM) total and scale scores by education among biological mother respondents with a diagnosed mental illness (N = 199) including depression (N = 191), Bipolar (N = 35), Schizophrenia (N = 1), and Dissociative Identity Disorder (N = 1) expressed as a percentage of the high-school group (Total = 12.0, SEM = 1.0; Internalizing = 2.8, SEM = 0.3; Externalizing = 4.8, SEM = 0.5; Attention = 4.5, SEM = 0.4). ^ap < .05 versus high-school, ^bp < .05 versus college. B) BPM by current annual income (in thousands) expressed as a percentage of the < 20 K/year group (Total = 11.7, SEM = 1.0; Internalizing = 3.1, SEM = 0.4; Externalizing = 4.6, SEM = 0.5; Attention = 4.0, SEM = 0.4). ^cp < .05 t-test versus <20 K/year, ^dp < .05 t-test versus 20 to 49 K/year.

Supplemental Table 1. Internal consistency of the Brief Problem Monitor and scales for biological mothers (N = 466), biological fathers (N = 26), adoptive (N = 54) and foster (N = 8) parents.

Supplemental Table 2. Internal consistency of the Brief Problem Monitor and scales by sex and age (children = 6 to 11, adolescents = 12 to 18).

Supplemental Table 3. High (T₅₀ > 65) Brief Problem Monitor total and scale scores by education and income (thousands/year) of the biological mother. ^Hp < .05 or ^HHp < .005 versus high-school; ^Cp < .05 versus college; ^Lp < .05 or ^LLp < .005 versus incomes less than 20K/year; ^Tp < .05 versus 20–49 K/year.

Supplemental Table 4. Brief Problem Monitor total and scale scores (±SEM) by ethnicity of the child. High = T₅₀ score ≥ 65. Other includes Black, Asian, Native Hawaiian, Pacific Islander, American Indian, Alaska Native, and other.

Appendix A. Brief Problem Monitor Parent Form (BPM-P) for ages 6–18

Appendix B. Consent and online questionnaire including maternal and child demographic items