Validation of the French Version of Conners’ Parent Rating Scale–Revised,Short Form in ADHD-Diagnosed Children and Comparison With Control Children

Abstract

Objective: The objective of this study was to examine measurement invariance and discriminant validity of the French Lausanne version (FLV) of the Conners’ Parent Rating Scale–Revised, Short Form (CPRS-R:S) and assess its convergent validity against the ADHD Symptoms Rating Scale (ADHD-SRS) and the Child Behavior Checklist (CBCL). Method: A confirmatory factor analysis and Tobit models were used in 108 ADHD children (aged 6-17) vs. 794 controls (aged 9-15) and score correlations were examined between FLV and ADHD-SRS then CBCL. Results: The factorial structure and reliability of the FLV is confirmed in ADHD children. The FLV showed configural invariance, metric invariance, and scalar invariance. FLV scores were significantly higher in ADHD than in control children in all three dimensions (p < .001). There were strong correlations between FLV Hyperactivity and Cognitive problems/Inattention scores and ADHD-SRS scores of Impulsivity/Hyperactivity (r = .90) and Inattention (r = .68) and also strong correlations between FLV Opposition score and the CBCL scores of Aggressive behavior (r = .84) and Rule-breaking behavior (r = .66). Conclusion: The study brings support for validation of the FLV regarding invariance in ADHD children, discriminant validity, and convergent validity.

Keywords

ADHD child adolescent validation behavior rating scale Conners’ scale ADHD Symptoms Rating Scale Child Behavior Checklist

Introduction

ADHD seems to affect 1% to 20% of schoolchildren. Various analyses have estimated the disease’s global prevalence in children and adolescents at 5.29% (Polanczyk, de Lima, Horta, Biederman, & Rohde, 2007) to 7.1% (Willcutt, 2012).

According to the current international guidelines, diagnosing ADHD requires a clinical examination and the use of reference questionnaires (Committee on Quality Improvement, Subcommittee on Attention-Deficit/Hyperactivity Disorder, 2000; Subcommittee on Attention-Deficit/Hyperactivity Disorder, Steering Committee on Quality Improvement and Management, 2011) such as the Conners’ Parent Rating Scale (CPRS; Conners, 1997) or the ADHD Rating Scale (DuPaul, Power, Anastopoulos, & Ried, 1998). More recently, the revised CPRS (CPRS-R) proved to be an interesting complementary diagnostic tool for ADHD (Gianarris, Golden, & Greene, 2001).

Worldwide, various Conners’ scales underwent several translations and validation processes but, to our knowledge, most French versions of the short form of the CPRS-R (Conners’ Parent Rating Scale–Revised, Short Form [CPRS-R:S]) have no validation support. In 2006, a collaboration between Multi-Health Systems (New York, USA, and Toronto, Canada) and a team from Lausanne, Switzerland (Pierrehumbert, Bader, Thévoz, Kinal, & Halfon, 2006) carried out a translation/back-translation of the American CPRS-R:S (Conners, 1997). Ten years later, another French–Swiss team (Fumeaux et al., 2016; Fumeaux et al., 2020) confirmed that this French Lausanne version (FLV) (a) has a strong three-dimensional factorial structure (identical to that of the 1997 American version), (b) has a satisfactory dimension internal consistency, (c) has good item reliability, and (d) is invariant across sex and age.

The objective of the present work was to continue the validation process of the FLV CPRS-R:S in children aged 6 to 17 years completed by studying (a) the measurement invariance across ADHD-diagnosed children and control children, (b) the discriminant validity between ADHD-diagnosed children and controls, and (c) the convergent validity with other frequently used scales, namely, the ADHD Symptoms Rating Scale (ADHD-SRS) (Holland, Gimpel, & Merrell, 2001) and the Child Behavior Checklist (CBCL; Achenbach & Rescola, 2000, 2001).

Method

Sample Size

The present work is an observational cross-sectional study carried out on two groups: a new group of ADHD children and a previously studied group of nearly 800 control children. Setting to 80% the power of detecting a difference of 0.3 standard deviations in mean score between the two groups with a two-sided significance level of 5%, 100 children in the ADHD group were needed. The important sample size of the control group allowed an acceptable precision for the 95% confidence intervals (CIs) of score means by sex and age-class.

The Control Group

By the end of 2011, the participation in the study was proposed to the parents of 1,218 middle school children or adolescents of a private institution (Lyon, France). Each questionnaire pack included an explanatory sheet, a sheet for child and family data (to help describing the sample and control for major confounding factors), and the FLV CPRS-R:S.

At form collection, the response rate was 65%; 794 filled out questionnaires were eligible for various analyses. The respondents were the mothers (70.4%), the fathers (11.3%), both parents (17.7%), or another person caring for the child (0.6% grandparent, stepparent, or elder brother/sister). After exclusion of 55 forms (for item “sex” left blank), 739 forms were kept for discriminant analysis (Figure 1). These forms concerned 365 boys and 374 girls; their mean ± SD age was 12.7 ± 1.2 years (range: 9-17) and the age distribution was 34% aged 9 to 11, 64% aged 12 to 14, and 2% aged 15 to 17 years completed. The parents belonged to the well-educated middle class: 79% of the children had at least one parent with tertiary education (71% of the mothers and 66% of the fathers) and 87% of the mothers and 97% of the fathers were in full-time or part-time paid employment.

Figure 1.

Flowchart.

The ADHD Group

From end-2011 to mid-2012, the participation in the study was proposed to the parents of all children suspected of ADHD and admitted to the unit of child and adolescent neuropsychopathology of a university hospital (Hôpital neurologique et neurochirurgical Pierre Wertheimer, Bron, France). One condition for inclusion was a confirmed diagnosis of ADHD according to the Diagnostic and Statistical Manual of Mental Disorders (4th ed.; American Psychiatric Association, 1994) criteria after medical, psychopathological, and neuropsychological assessments. Another condition was the parents’ and child consents to participation after full information.

The pack of questionnaires included an explanatory sheet, a sheet for child and family data, the FLV CPRS-R:S, the ADHD-SRS, and the CBCL. The response rate was 65% (126 filled out questionnaires out of 194 proposed). The responders were the mothers (70.9%), the fathers (4.5%), or both parents (24.5%). After exclusion of 18 records (17 because of unconfirmed diagnosis and one because of age >17), 108 ADHD children (87 boys and 21 girls) were kept for analysis (Figure 1). Their mean age ± SD was 12.0 ± 3.3 years (range: 6-17). On average, this ADHD group was 8 months younger than the control group (p = .019). The age distribution was 22% aged 6 to 8, 30% aged 9 to 11, 36% aged 12 to 14, and 22% aged 15 to 17 years completed. In this group, 80.6% were boys (vs. 49.4% in the control children group, p < .001) and 58% of the children had at least one parent with tertiary education (58% of the mothers and 52% of the fathers), which is lower than in the control children (p < .001) and 78% of the mothers and 94% of the fathers were in full-time or part-time paid employment (respectively, slightly lower and similar percentages to those of the control group, p = .013 and p = .19).

The Data

The CPRS-R:S data were the scores given to the 18 items of all three dimensions (six items per dimension) of the FLV CPRS-R:S (Conners, 1997). These dimensions are (a) Oppositional: Items 2, 6, 11, 16, 20, and 24; (b) Cognitive problems/Inattention: Items 3, 8, 12, 17, 21, and 25; and (c) Hyperactivity: Items 4, 9, 14, 18, 22, and 26. Each item is rated on a 4-point Likert-type scale: 0 for “Pas du tout vrai” (Not true at all), 1 for “Un peu vrai” (Just a little true), 2 for “Assez vrai” (Pretty much true), and 3 for “Très vrai” (Very much true).

The ADHD-SRS data were the scores given to the 18 items of dimensions Inattention and Hyperactivity/Impulsivity (nine items per dimension). Each item is rated on a 4-point Likert-type scale according to the symptom frequency: 0 for “Jamais ou rarement” (Never or rarely), 1 for “Parfois” (Sometimes), 2 for “Souvent” (Often), and 3 for “Très souvent” (Very often).

The CBCL data were the scores given to the 112 items of six dimensions: Depression-Anxiety, Social problems, Thought problems, Attention problems, Rule-breaking behavior, and Aggressive behavior. Each item is rated on a 3-point Likert-type scale: 0 for “Pas vrai (à votre connaissance)” (Not true as far as you know), 1 for “À peu près vrai ou parfois vrai” (Somewhat or sometimes true), and 2 for “Très vrai ou souvent vrai” (Very true or often true). The dimension of particular interest was “Attention problems” (11 items).

Statistical Methods

Confirmation of the factorial structure and reliability of the FLV in ADHD children

To confirm the factorial structure, we used a confirmatory factor analysis (CFA, Bollen, 1989) as the analysis used on the control group (Fumeaux et al., 2016). Starting with a baseline model (three correlated dimensions with six items in each and no residual correlations between items), other models were used to improve the fit and check whether the models published for the controls apply to ADHD children. In the controls, the final model was the baseline model plus two residual correlations between Items 4 and 18 of dimension Hyperactivity and between Items 6 and 20 of dimension Oppositional.

Several indices were used to assess the fit of the CFA models: (a) the comparative fit index (CFI) and the Tucker–Lewis index (TLI) with “good fit” if > 0.95; (b) the root mean square error of approximation (RMSEA) and its 90% CI with “close fit” if < 0.05 and “fair fit” in the 0.05 to 0.08 range (Browne & Cudeck, 1992); (c) the weighted root mean square residual (WRMR), with “good fit” if ≤ 1.0; and (d) the modification indices (MIs) of model improvement, a high MI indicating an important improvement.

Invariance of the FLV CPRS-R:S between ADHD children and controls

By invariance, it is meant that individuals with same score on a scale have same status on the construct of this scale whatever their other characteristics. Assessing invariance is thus measuring the degree to which a scale measures the same construct in different groups.

The assessment of invariance of the FLV CPRS-R:S between ADHD children and controls used the method that assessed previously the invariance across sex and age-class (Fumeaux et al., 2020). This method is based on building CFA models to estimate the regression coefficient (or factor loadings) of each item on the latent variable it is meant to reflect.

Briefly, three forms of invariance were considered: configural, metric, and scalar (Meredith & Teresi, 2006; Teresi, 2006). Configural invariance is supported when the CFA models have the same number of dimensions in the two groups and when the same item reflects the same dimension in the two groups; that is, have a nonzero factor loading on this dimension and a zero factor loading on the other dimensions. Metric invariance is supported when, in addition, the item factor loadings are equal in the two groups. Scalar invariance is supported when, furthermore, the item thresholds are equal in the two groups (here, the number of thresholds is three = 4 rates − 1), which allows comparing dimension means or item means.

The invariance approach considers a series of three nested CFA models (configural, metric, and scalar) and tests differences between these models (configural vs. metric and metric vs. scalar). As these forms are nested, the assessment of each form assumes that the previous form is supported; this leads to consider that metric invariance is weak and scalar invariance is strong. Nevertheless, each form of invariance may be only partial (Byrne, Shavelson, & Muthén, 1989), that is, one or more items may not show invariance (either weak or strong). In this case, a closer examination of the factor loadings and thresholds is needed. Plotting the estimations of these parameters (with their CIs) helps showing the extents of their differences between groups.

Changes in fit indices between nested models are also reported; that is, ΔCFI and ΔRMSEA, although there is no consensus regarding their cutoff values (Putnick & Bornstein, 2016).

Discriminant validity

FLV CPRS-R:S scores were compared between ADHD children and controls. Descriptive statistics are displayed to allow comparisons with previously published data.

Score comparisons were adjusted for age (as continuous variable) and sex. In these adjusted comparisons, a Tobit model (Long, 1997; Tobin, 1958), or “censored regression model” —designed to estimate linear relationships between variables in case of left- or right-censoring in the dependent variable— was used to deal with the floor effect in the control group and a potential ceiling effect in the ADHD group in each dimension (score minimum 0, maximum 18). The mean scores of the ADHD-diagnosed children were also examined by dimension, sex, and age-class (6-8, 9-11, 12-14, and 15-17 years, as proposed by Conners, 1997). A further class 9 to 15 years was also considered for comparisons with previously published scores on the controls.

Convergent validity

Pearson correlation with z transformation of Fisher was used to evaluate the convergent validity of the FLV CPRS-R:S in the ADHD group. A strong correlation (correlation coefficient > .6) is expected (a) between the score of Hyperactivity in the CPRS-R:S and the score of Impulsivity/Hyperactivity score of the ADHD-SRS, and (b) between the score of Attention problems in the CBCL, the Inattention score of the ADHD-SRS, and the Cognitive problems/Inattention score of the CPRS-R:S.

Software programs

The CFA models and the CIs of their parameters were estimated using Mplus, version 7.4 (Muthén & Muthén, 1998-2012). R software version 3.2.2 (http://www.r-project.org/) and SAS® version 9.4 were used for all other analyses. All tests were two-tailed and p < .05 was considered for statistical significance.

Results

Factorial Structure and Reliability of the FLV CPRS-R:S in ADHD Children

In the ADHD group, the FLV CPRS-R:S was fitted to distinct CFA models with same specifications (Table 1). The baseline model (Model 1) had a good fit as per the CFI, the TLI (both > 0.95), and the WRMR (<1) but had a poor fit as per the RMSEA (0.083; 90% CI: [0.064, 0.100], hypothesis RMSEA ≤ .05 rejected). A residual correlation between Items 4 and 18 of dimension Hyperactivity was suggested because of a high MI. After adding this residual correlation, another residual correlation between Items 6 and 20 of dimension Oppositional was suggested. This led successively to Models 2 and 3 (Table 1) with improved fit at each step, although the improvement induced by Model 3 was minor as per the statistical criteria. The final model for ADHD children was Model 3; that is, the baseline model + residual correlations between Items 4 and 18 and between Items 6 and 20 (residual correlations: .661 and .551, respectively).

Table 1.

Model Fit Criteria in ADHD Children.

Model and characteristics	χ²	df	CFI	TLI	RMSEA		WRMR	Model improvement
Model and characteristics	χ²	df	CFI	TLI	Value [90% CI]	p value	WRMR	Parameter to add	MI	CS-EPC
Model 1: Baseline model	229.310	132	0.958	0.951	0.083 [0.064, 0.100]	.003	0.942	Residual correlation between Items 4 and 18	33.9	1.385
Model 2: Model 1 + residual correlation between Items 4 and 18	199.325	131	0.970	0.965	0.069 [0.049, 0.088]	.057	0.839	Residual correlation between Items 6 and 20	11.1	−0.227
Model 3: Model 2 + residual correlation between Items 6 and 20	193.686	130	0.972	0.967	0.067 [0.046, 0.086]	.082	0.820	Item 8 reflecting Hyperactivity	11.2	−0.227

Note. df = degrees of freedom, CFI = comparative fit index; TLI = Tucker–Lewis index; RMSEA = root mean square error of approximation; CI = confidence interval; WRMR = weighted root mean square residual; MI = modification index; CS-EPC = completely standardized expected parameter change.

Invariance of the FLV CPRS-R:S Between ADHD and Control Children

In the three models (i.e., Models 1-3), the factor loadings, the correlation coefficients between dimensions, and the residual correlations were all statistically significantly different from 0. Moreover, all the factor loadings were positive. These results demonstrate that the whole scale had a configural invariance across ADHD children and controls.

Figure 2 shows comparisons of factor loadings per dimension and item. In only one item (Item 6 of dimension Oppositional), the 95% CIs of the factor loadings did not overlap (1.257 [1.043, 1.471] in the ADHD group vs. 0.900 [0.813, 0.946] in the control group). Thus, the two factor loadings were significantly different. However, the difference in model adequacy to the data between the metric model (factor loadings constrained to be equal across groups) and the configural model (free factor loadings) was not statistically significant (Table 2, p =.06). This agrees with the hypothesis of equal factor loadings between the two groups. Furthermore, the fit indices did not change or changed slightly (ΔCFI <0.001, ΔRMSEA = −0.001). These results demonstrate that the whole scale had a metric invariance across the two groups.

Figure 2.

Factor loadings of the items of the FLV CPRS-R:S (with their 95% confidence intervals) by children group and by dimension.

Table 2.

Model Fit Indices for Evaluating Invariance.

Model	χ²	df	CFI	TLI	RMSEA		WRMR	Invariance testing
Model	χ²	df	CFI	TLI	Value [90% CI]	p value	WRMR	Models compared	χ²	df	p value
Configural	507.988	260	0.984	0.981	0.046 [0.040, 0.052]	.865	1.335	—
Metric	523.016	275	0.984	0.982	0.045 [0.039, 0.051]	.932	1.372	Metric vs. Configural	24.034	15	.0645
Scalar	571.894	308	0.983	0.983	0.044 [0.038, 0.049]	.972	1.463	Scalar vs. Metric	64.282	33	< .001

The study of the scalar invariance is illustrated by Figure 3. Only one threshold showed no invariance across groups (Threshold 2 of Item 21 of dimension Cognitive problems/Inattention). Dimensions Oppositional and Hyperactivity showed a scalar invariance (at all thresholds), whereas dimension Cognitive problems/Inattention showed only a partial scalar invariance. The difference in model adequacy to the data between the scalar model (thresholds constrained to be equal across groups) and the metric model (free thresholds) was statistically significant (p < .001) but the fit indices were not significantly different (ΔCFI and ΔRMSEA = −0.001). These results indicate that, strictly speaking, the whole scale had a partial scalar invariance across the two groups.

Figure 3.

Threshold values of the FLV CPRS-R:S items (with their 95% confidence intervals) by children group and by dimension.

Discriminant Validity of the FLV CPRS-R:S

The scores of the FLV CPRS-R:S per dimension and group are summarized in Table 3 and the mean (SD) per dimension, age-class, and sex are displayed in Table 4. The mean and median scores in the control group were ≤3 in the three dimensions. As expected, the scores (adjusted for age and sex) were higher in ADHD children than in the controls (p < .001 in each dimension). The adjusted mean difference (SD) was 6.34 (4.98) for Oppositional, 8.63 (4.52) for Hyperactivity, and 11.12 (5.37) for Cognitive problems/Inattention. The corresponding effect sizes were 1.28, 1.91, and 2.07, respectively.

Table 3.

Distribution of the Scores of the French Lausanne version of CPRS-R:S Per Dimension in ADHD and Control Children.

Dimension and group	n	M	SD	Q1	Median	Q3	Maximum	p value
Oppositional								< .001
ADHD	108	8.15	5.32	4	8	12	18
Control	794	2.72	3.38	0	2	4	18
Hyperactivity								< .001
ADHD	108	8.39	4.84	4	9	12	18
Control	794	1.21	2.27	0	0	1	15
Cognitive problems/inattention								< .001
ADHD	108	13.12	4.27	10	14	17	18
Control	794	2.93	3.88	0	1	4	18

Note. Q1 and Q3: 25th and 75th percentiles of the distribution. The minimum score of any dimension is zero and its maximum 18. The p value is for the test of ADHD children versus control children using a Tobit model adjusted on age and sex. CPRS-R:S = Conners’ Parent Rating Scale–Revised: Short Form.

Table 4.

Mean (SD) Scores of the French Lausanne version of CPRS-R:S per Dimension and Subgroup in ADHD Children.

Subgroup	n	Oppositional	Hyperactivity	Cognitive problems/inattention
Boys	87	8.6 (5.3)	8.8 (4.7)	13.4 (4.2)
Girls	21	6.4 (4.9)	6.5 (4.9)	12.0 (4.4)
6-8 years	24	8.1 (5.3)	11.5 (4.0)	12.6 (3.9)
9-11 years	32	7.0 (4.9)	8.1 (5.0)	13.7 (4.0)
12-14 years	28	8.8 (5.5)	8.2 (4.1)	13.6 (4.1)
15-17 years	24	9.0 (5.8)	6.0 (4.9)	12.4 (5.1)
12-15 years	39	8.4 (5.2)	7.3 (4.2)	13.3 (4.5)

Note. The minimum score of any dimension is zero and its maximum 18. In the last line, children aged 15 years are grouped with those aged 12 to 14 years to allow comparisons with previously published results on the controls. CPRS-R:S = Conners’ Parent Rating Scale–Revised: Short Form.

Convergent Validity of the FLV CPRS-R:S

A very strong correlation was shown between the score of Hyperactivity of the FLV CPRS-R:S and the score of Impulsivity/Hyperactivity of the ADHD-SRS (r = .90; Table 5). The score of Cognitive problems/Inattention of the FLV CPRS-R:S was more correlated with the score of Inattention of the ADHD-SRS than with the score of Attention problems of the CBCL (r = .68 vs. .41). The score of Opposition of the FLV CPRS-R:S was very strongly correlated with the score of Aggressive behavior of the CBCL (r = .84) and strongly correlated with the score of Rule-breaking behavior of the CBCL (r = .66) and the score of Impulsivity/Hyperactivity of the ADHD-SRS (r = .62). As expected, the other estimated correlations between various scores were moderately strong to very weak (< .6; Table 5).

Table 5.

Spearman Correlation Coefficients Between Various Scale Dimensions (Convergent Validity).

Scale and dimensions	FLV CPRS-R:S dimensions
Scale and dimensions	Oppositional	Cognitive problems/Inattention	Hyperactivity
ADHD
Inattention	.47	.68	.44
Hyperactivity–Impulsivity	.62	.28	.90
CBCL
Depression–Anxiety	.44	.20	.34
Social problems	.35	.13	.44
Thought problems	.37	.18	.41
Attention problems	.52	.41	.59
Rule-breaking behavior	.66	.27	.36
Aggressive behavior	.84	.30	.56

Note. FLV = French Lausanne version; CPRS-R:S = Conners’ Parent Rating Scale–Revised: Short Form; CBCL = Child Behavior Checklist.

Discussion

This work continues the process of validation of the FLV CPRS-R:S. A first step has demonstrated the validity of the three-dimensional factorial structure, the dimension internal consistency, and the item reliabilities in a sample of control children (Fumeaux et al., 2016). A second step checked the scale invariance across sex and age-classes (Fumeaux et al., 2020). The present study is a third step of this validation that aims to offer French-speaking health professionals and parents a reliable tool to assist diagnosis and, ultimately, be used in clinical trials. We assess here the FLV measurement invariance across ADHD-diagnosed and control children, the FLV discriminant validity between ADHD-diagnosed and control children, and the FLV convergent validity with the ADHD-SRS and the CBCL in ADHD-diagnosed children.

The factorial structure of the FLV CPRS-R:S was confirmed in ADHD-diagnosed children. The model already established in the control group proved to be well fitted in ADHD children; that is, here too, the baseline model required two residual correlations between Items 4 and 18 of dimension Hyperactivity and between Items 6 and 20 of dimension Oppositional. With the present study groups, the fit was not significantly better with two than with one residual correlation between Items 4 and 18. However, to study the invariance, we chose to compare the two groups using the same model with two correlations because this model was already validated on the much larger control group (Fumeaux et al., 2016; Fumeaux et al., 2020).

We considered here three forms of invariance. For dimensions Oppositional and Hyperactivity, we found a scalar invariance but for dimension Cognitive problems/Inattention, only a partial scalar invariance was found. This was because one threshold of one item (threshold 2 of item 21) was different between the two groups. However, because the other thresholds for this item and all the thresholds of all other items were equal between the two groups and because the difference between the metric model (factor loadings constrained to be equal) and the configural model (free factor loadings) was not statistically significant, a scalar invariance may be also acknowledged for dimension Cognitive problems/Inattention. Furthermore, configural invariance being demonstrated, meaningful comparisons of group means can be undertaken when most factor loadings per dimension show scalar invariance across groups. Given the present invariance results, score comparisons between ADHD and control children of same sex and age should be meaningful.

After adjustment for sex and age, the discriminant validity of the FLV was confirmed; this scale can thus be used as complementary diagnostic tool for ADHD. Indeed, in a previous work (Fumeaux et al., 2020) the mean scores were higher in ADHD-diagnosed children than in the control children.

In ADHD-diagnosed children, the convergent validity of the FLV CPRS-R:S was demonstrated (a) by the very strong correlation between the FLV Hyperactivity score and the Impulsivity/Hyperactivity score of the ADHD-SRS; (b) by the moderately strong correlation between the FLV Cognitive problems/Inattention score and the Attention problems score of CBCL; and by the strong correlation between FLV Cognitive problems/Inattention score and the Inattention score of ADHD-SRS.

Limitations

One The main limitation of the present study was that sex and age distributions were not “comparable” between ADHD and control children. Precisely, the proportion of boys was higher among ADHD children than among the controls (80.6% vs. 49.4%). However, this corresponds to the sex ratio usually observed in ADHD (3:1 to 4:1). Also, the age distribution was larger in ADHD children than in the controls because there were more children aged <9 or >15 years in the former group.

Clinical Implications

The completion and favorable results of this third step of validation increase the physician’s confidence in the FLV CPRS-R:S (all dimensions) regarding its use for comparisons between typically developing and ADHD-suspected children in similar populations; for example, its use in international clinical trials.

Conclusion

The FLV CPRS-R:S, already validated in a control group regarding its factorial structure, internal consistency, reliability, and invariance across sex and age, is now validated regarding its invariance across ADHD-diagnosed and control children, its convergent validity, and its discriminant validity between ADHD-diagnosed children and control children. This additional validation step confers the FLV CPRS-R:S a higher value in French-speaking countries for comparing normal with ADHD-suspected children (all dimensions) in similar populations. It also supports further its use in international clinical trials.

Footnotes

Acknowledgements

The authors thank Charles de Foucauld School for providing all possible help and all the children and adolescents and their families for their participation. The authors also thank Dr. Michel Bader for initiating the research project and providing the French versions of the questionnaires and especially thank Prof. René Ecochard for valuable advice throughout the conduct of this study.

Authors’ Note

In all cases, participation was voluntary. Parent or child refusal to participate was an exclusion criteria as well as the return of a blank questionnaire. For confidentiality, there were no names on the questionnaires and the participants had the assurance of that all data will be used in aggregate form only. This study was carried out with agreements from (a) the Comité de Protection des Personnes Lyon Sud Est II, France (obtained on January 5, 2012); (b) the Comité Consultatif sur le Traitement de l’Information en matière de Recherche dans le domaine de la Santé, France (Agreement Nr. 12.185bis issued on September 6, 2012; and, (c) the Commission Nationale de l’Informatique et des Libertés, France (Agreement Nr. DR-2012-524, issued on November 7, 2012).

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article from: (a) Fonds de Perfectionnement du Centre Hospitalier Universitaire Vaudois (CHUV); (b) Société Académique Vaudoise; (c) Fondation d’Entreprise Laboratoire Urgo; and (d) Shire Pharmaceuticals Group.

Author Biographies

Pierre Fumeaux, MD, is specialized in child and adolescent psychiatry in private practice in Lausanne (Switzerland). He is a clinician and researcher, working in the field of ADHD, High Intellectual Potential, and learning disabilities.

Catherine Mercier, PhD, is a biostatistician at the Service de Biostatistique-Bioinformatique, Hospices Civils de Lyon, Lyon, France. Her research focuses on biostatistical methods for clinical trials, biomarker discovery, and questionnaire validation.

Sylvain Roche, MSc, is a biostatistician at the Service de Biostatistique-Bioinformatique, Hospices Civils de Lyon, Lyon, France. His research focuses on latent variable modeling in epidemiology and psychometrics.

Jean Iwaz, PhD, is a scientific and medical editor at Service de Biostatistique-Bioinformatique, Hospices Civils de Lyon, Lyon, France. He has over 25 years’ experience in manuscript analysis and critical comment.

Philippe Stéphan, MD, is a clinician and researcher at the University of Lausanne Child and Adolescent Psychiatry Department (Switzerland).

Olivier Revol, MD, is the head of the Service de neuropsychiatrie de l’enfant de l’adolescent. He is also involved in the Centre de référence troubles spécifiques du langage et des apprentissages, Hospices Civils de Lyon, France. His main research interests are focused on ADHD, High Intellectual Potential, learning disabilities, and emotional disorders of the child.

References

Achenbach

T. M.

Rescola

L. A.

(2000). Manual for the ASEBA preschool forms and profiles. Burlington: Research Center for Children, Youth, and Families, University of Vermont. Available from www.aseba.org

Achenbach

T. M.

Rescola

L. A.

(2001). Manual for the ASEBA school-age forms and profiles. Burlington: Research Center for Children, Youth, and Families, University of Vermont. Available from www.aseba.org

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author.

Bollen

K. A.

(1989). Structural equations with latent variables. New York, NY: John Wiley.

Browne

M. W.

Cudeck

(1992). Alternative ways of assessing model fit. Sociological Methods & Research, 21, 230-258.

Byrne

B. M.

Shavelson

R. J.

Muthén

(1989). Testing for the equivalence of factor covariance and mean structures: The issue of partial measurement invariance. Psychological Bulletin, 105, 456-466.

Committee on Quality Improvement, Subcommittee on Attention-Deficit/Hyperactivity Disorder. (2000). Clinical practice guideline: Diagnosis and evaluation of the child with attention-deficit/hyperactivity disorder. Pediatrics, 105, 1158-1170.

Conners

C. K.

(1997). Conners’ Rating Scales-Revised Technical Manual. North Tonawanda, NY: Multi-Health Systems. Available from http://www.mhs.com

DuPaul

G. J.

Power

T. J.

Anastopoulos

A. D.

Ried

(1998). ADHD Rating Scale–IV: Checklists, norms, and clinical interpretations. New York, NY: Guilford Press.

10.

Fumeaux

Mercier

Roche

Iwaz

Bader

Stéphan

Ecochard

Revol

(2016). Validation of the French version of Conners’ Parent Rating Scale Revised, Short Version: Factorial structure and reliability. The Canadian Journal of Psychiatry, 61, 236-242.

11.

Fumeaux

Roche

Mercier

Iwaz

Bader

Stéphan

Ecochard

Revol

(2020). Validation of the French version of Conners’ Parent Rating Scale–Revised, Short Version (CPRS-R:S): Scale measurement invariance by sex and age. Journal of Attention Disorders, 24, 1693-1700. doi:10.1177/1087054717696767

12.

Gianarris

W. J.

Golden

C. J.

Greene

(2001). The Conners’ Parent Rating Scales: A critical review of the literature. Clinical Psychology Review, 21, 1061-1069.

13.

Holland

M. I.

Gimpel

G. A.

Merrell

K. W.

(2001). ADHD Symptoms Rating Scale Manual. Wilmington, DE: Wide Range. Retrieved from www4.parinc.com/Products/Product.aspx?ProductID=ADHD-SRS

14.

Long

J. S.

(1997). Regression models for categorical and limited dependent variables. Thousand Oaks, CA: SAGE.

15.

Meredith

Teresi

J. A.

(2006). An essay on measurement and factorial invariance. Medical Care, 44, S69-S77.

16.

Muthén

L. K.

Muthén

B. O.

(1998-2012). Mplus user’s guide (7th ed.). Los Angeles, CA: Author. Retrieved from http://www.statmodel.com/download/usersguide/Mplus%20user%20guide%20Ver_7_r6_web.pdf

17.

Pierrehumbert

Bader

Thévoz

Kinal

Halfon

(2006). Hyperactivity and attention problems in a Swiss sample of school-aged children: Effects of school achievement, child gender, and informants. Journal of Attention Disorders, 10, 65-76.

18.

Polanczyk

de Lima

M. S.

Horta

B. L.

Biederman

Rohde

L. A.

(2007). The worldwide prevalence of ADHD: A systematic review and metaregression analysis. American Journal of Psychiatry, 164, 942-948.

19.

Putnick

D. L.

Bornstein

M. H.

(2016). Measurement invariance conventions and reporting: The state of the art and future directions for psychological research. Development Review, 41, 71-90.

20.

Subcommittee on Attention-Deficit/Hyperactivity Disorder, Steering Committee on Quality Improvement and Management. (2011). ADHD: Clinical practice guideline for the diagnosis, evaluation, and treatment of attention-deficit/hyperactivity disorder in children and adolescents. Pediatrics, 128, 1-16.

21.

Teresi

J. A.

(2006). Overview of quantitative measurement methods. Equivalence, invariance, and differential item functioning in health applications. Medical Care, 44, S39-S49.

22.

Tobin

(1958). Estimation of relationships for limited dependent variables. Econometrica, 26, 24-36.

23.

Willcutt

E. G.

(2012). The prevalence of DSM-IV attention-deficit/hyperactivity disorder: A meta-analytic review. Neurotherapeutics, 9, 490-499.