The Screen for Child Anxiety Related Emotional Disorders (SCARED) is a 41-item rating scale that measures anxiety problems in youth. The SCARED has versions for multiple informants and strong psychometric support. A five-item version (SCARED-5) has been proposed as a brief anxiety screening tool, but limited research has examined its psychometric properties. This study evaluated the SCARED-5 in two samples of youth (Study 1: N = 109 youth, Mage = 12 years; Study 2: N = 175 youth, Mage = 8 years). Results supported the one-factor structure and internal consistency reliability of the SCARED-5 across informants. Total scores from the SCARED-5 were significantly associated with total scores from the full SCARED. Mother and child ratings on the SCARED-5 successfully distinguished anxious and non-anxious youth, but father ratings did not. Receiver operating characteristic analyses supported a clinical cutoff score of 2 on the SCARED-5 for screening purposes. These findings provide preliminary validation for the SCARED-5 as an efficient anxiety screening tool in clinical and research settings. Future research should examine the SCARED-5 in more diverse samples and validate informant-specific cutoff scores, particularly for father ratings.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Anxiety disorders are one of the most common mental health disorders in youth, affecting an estimated 31% of children and adolescents at some point during development (Merikangas et al., 2010). Anxiety disorders can have a significant negative impact on a child’s life, interfering with social, emotional, and academic functioning (Bittner et al., 2007; Bushnell et al., 2020). Early identification and intervention are essential to improve outcomes, as untreated anxiety often persists into adulthood and increases the risk for other psychiatric conditions (Bushnell et al., 2020; Poulton et al., 2023). Standardized assessment tools play a key role in this early identification. The Screen for Child Anxiety-Related Emotional Disorders (SCARED) is a 41-item rating scale that assesses a range of anxiety disorders in youth ages 5 to 19 (Birmaher et al., 1999; Birmaher et al., 1997; Sequeira et al., 2019). The SCARED is widely used in clinical settings and research studies. The popularity of the SCARED is likely because it is in the public domain (i.e., freely available for use) and because it has strong psychometric support in clinical and community samples in the U.S. and around the world (Behrens et al., 2019; Hale et al., 2011).
The SCARED-5 is a five-item version derived from the SCARED-41, developed for efficient anxiety screening in youth. When proposing the SCARED-5, Birmaher et al. (1999) showed that the SCARED-5 had similar psychometric properties to the full SCARED, with a sensitivity of 74% and a specificity of 73%. However, no subsequent research has evaluated the SCARED-5. Further validation is important for two reasons. First, the SCARED-5 is already used in research and clinical practice despite the limited data on its psychometric properties (Chavira et al., 2004; Gjerde et al., 2023; Stepp et al., 2010; Waschbusch et al., 2020). Second, confirming its psychometric soundness is essential because there is a significant need for validated brief screening tools for anxiety disorders in youth mental health (U. S. Preventive Services Task Force et al., 2022; Viswanathan et al., 2022). Practical brief screening tools are particularly valuable in time-constrained settings, like primary care and schools, where they can facilitate early identification and referral. Establishing the SCARED-5’s properties will determine its suitability to meet this need and support its appropriate use as an important alternative to the full SCARED, which is more comprehensive but takes longer to complete. Developing both brief and comprehensive measures is an important step toward effective and comprehensive care of mental health disorders, allowing clinicians to select the appropriate depth of assessment based on setting constraints or clinical needs (Rammstedt & Beierlein, 2014; Ziegler et al., 2014).
In both screening and clinical evaluations, there is value in incorporating multiple informants, such as mothers, fathers, and the youth themselves (von der Embse et al., 2019). Different informants provide unique perspectives on the youth’s mental health; parents may observe behavioral manifestations of anxiety across settings, while youth can report on internal experiences that may not be visible to others. Previous research on anxiety assessment has consistently documented informant discrepancies between parents and children, with varying patterns of agreement depending on the specific symptoms, contexts, and developmental stages being assessed (Becker-Haimes et al., 2018; Comer & Kendall, 2004), and mothers often report more symptoms than fathers (Jansen et al., 2017; Krain & Kendall, 2000; Schroeder et al., 2010; Treutler & Epkins, 2003). Incorporating these different perspectives can provide a more complete picture of the youth’s adjustment (De Los Reyes & Epkins, 2023). Accordingly, examining how the SCARED-5 performs across different informants is essential for determining its optimal clinical use.
This study evaluated the psychometric properties of the SCARED-5, including its factor structure, reliability, convergent validity, and diagnostic accuracy, in two complementary samples. We specifically examined the measure’s performance across mother, father, and child reports, addressing an important gap in the literature. The first sample was from a previously published study (Petrovic-Dovat et al., 2016) which examined anxiety in youth recruited from different clinics. The second sample included youth receiving services in an outpatient mental health assessment clinic. Together, these samples allow us to address the following research questions: (1) Does the SCARED-5 demonstrate a robust factor structure and internal consistency? (2) How well does the SCARED-5 correlate with the full SCARED-41? (3) Can the SCARED-5 effectively distinguish between anxious and non-anxious youth? and (4) What cutoff score on the SCARED-5 optimizes screening accuracy? The method and results for each study are presented, followed by a general discussion integrating findings across both studies.
Study 1
Method
Participants
Participants were 109 youth (53.2% boys) ages 8 to 16 years (M = 12.33, SD = 2.37). Demographic information was collected via parent report. Parents reported their child’s race using a “check all that apply” format. The reported racial distribution was as follows: Asian (n = 6, 5.5%), Black or African American (n = 11, 10.1%), Native American or Alaska Native (n = 1, 0.9%), White (n = 98, 89.9%), and Other Race (n = 1, 0.9%). Seven participants (6.4%) endorsed more than one race. Regarding ethnicity, 9 participants identified as Hispanic or Latino (8.4% of the 107 reporting ethnicity; ethnicity data were missing for two participants). Data completeness for the primary measures was high; most participants (95.4%) had both child self-ratings and mother ratings, but four participants (3.7%) had mother ratings but not child ratings, and one participant (0.9%) had child ratings but not mother ratings.
Procedure
The institutional review board approved this cross-sectional study, where parents or legal guardians provided written informed consent, and child participants gave verbal assent. This study was not preregistered. Youth were recruited from three outpatient clinics at an academic medical center: an allergy and immunology clinic (n = 35; 32.1%), a general pediatrics clinic (n = 39; 35.8%), and an anxiety disorders clinic (n = 35; 32.1%). A study team member approached participants and their mothers while waiting for a medical visit. The study was described, and individuals were asked to participate. Children and mothers who agreed to participate completed the ratings described below. Mothers were instructed to assist children younger than 14 with filling out ratings as needed. The number of individuals who declined participation was not recorded; however, nearly all families who were approached agreed to participate, with lack of time cited as the most common reason for refusal.
Following data collection, participants were classified into anxious (n = 43) and non-anxious (n = 66) groups based on diagnostic information derived from multiple sources. For participants recruited from the specialty anxiety clinic (n = 35 of the 43 in the final anxious group), anxiety disorder diagnoses had been previously confirmed via clinical assessment by child psychiatrists or psychologists based on Diagnostic and Statistical Manual criteria. For all participants, information regarding history of anxiety or obsessive-compulsive disorders was also gathered from parent and child report using the Anxiety Disorder Questionnaire (described below). All diagnostic information (whether from clinic designation or screen reports) was corroborated through a review of participants’ electronic medical records by a study team member. The medical record served as the definitive source; participants were assigned to the anxious group only if a relevant anxiety disorder diagnosis was documented in the record or confirmed by the original clinical assessment for the anxiety clinic subgroup.
Measures
Screen for child anxiety related disorders (SCARED)
The SCARED assesses symptoms of anxiety using 41 items that are rated using 0 (Not True or Hardly Ever True), 1 (Somewhat True or Sometimes True), or 2 (Very True or Often True) as the response format. A total score for the full SCARED (SCARED-41) was computed by summing all 41 items, and a total score for the screening version (SCARED-5) was computed by summing items 24, 25, 28, 36, and 41. Proposed cutoff scores for identifying anxiety are 25 or higher for the SCARED-41 and 3 or higher for the SCARED-5 (Birmaher et al., 1999), although the cutoff from the SCARED-5 will be a focus of this study. Item-level descriptive statistics for the SCARED-41 are shown in an online supplement table S1.
Anxiety disorder questionnaire (ADQ)
The ADQ was an investigator-designed questionnaire developed for the original study to collect background information regarding anxiety disorders (Petrovic-Dovat et al., 2016). The measure included sections for child report (with instructions for parents to assist children younger than 14 years) and parent report. It inquired about the history of diagnosis and treatment (including medication and therapy) for anxiety disorders in the child and the parent’s history of these conditions. The format primarily used Yes/No questions with brief open-ended follow-ups.
Statistical analyses
Data were analyzed using SPSS version 28.0.1, MPLUS version 8.9 (Muthén & Muthén, 1998–2023), and R version 4.5.0 (R Core Team, 2025) with the pROC (Robin et al., 2011) and OptimalCutpoints (López-Ratón et al., 2014) packages. The following psychometric evaluation steps were taken for each study where applicable:
1.
As a preliminary step for Study 1, potential differences across the three recruitment sites were examined for participant age and sex distribution (using ANOVA and chi-square tests, respectively) and mean SCARED scores (using ANOVAs with Tukey post-hoc tests) to better understand the sample characteristics given the mixed recruitment sources.
2.
Item-level descriptive statistics (frequencies, means, standard deviations) for the SCARED-5 items were computed to examine item endorsement patterns, variability, and basic psychometric characteristics.
3.
To evaluate the internal structure validity of the SCARED-5 (i.e., whether the items measure a single underlying construct as intended), confirmatory factor analyses (CFAs) were computed separately for each informant (mother, child, father). A single-factor model was tested using the WLSMV estimator in MPLUS, appropriate for categorical item data. Model fit, indicating how well the proposed one-factor structure represented the observed data, was assessed using standard indices and cutoffs for acceptable fit: Root Mean Square Error of Approximation (RMSEA) < 0.08, Comparative Fit Index (CFI) > 0.95, and Standardized Root Mean Square Residual (SRMR) < 0.08 (Asparouhov & Muthén, 2018; Hu & Bentler, 1999). Wolf et al. (2013) demonstrated that for simple unidimensional models with 4-8 indicators, sample sizes of 90-190 were adequate for models with loadings of 0.50-0.65. The present samples fall within this empirically supported range. Following CFA, the internal consistency reliability of the factor, representing the degree to which the items consistently measure the same construct, was estimated using Coefficient H and Omega (McNeish, 2018), modern indices often preferred over Cronbach’s alpha for factor-based scales.
4.
To assess convergent validity, Pearson correlations were computed examining the association between scores on the brief SCARED-5 and the established full-length SCARED-41. Strong positive correlations would support that both scales measure the same underlying anxiety construct. Correlations were also computed between SCARED scores and participant age and sex (coded 0 = female, 1 = male) for descriptive purposes. Race and ethnicity variables were not included in correlation analyses due to insufficient sample sizes within several non-White racial and ethnic groups in these samples, precluding reliable analysis.
5.
To compare the overall score levels generated by the two versions, paired samples t-tests were computed between mean item scores (total score divided by number of items, placing scores on the same 0-2 metric) on the SCARED-5 versus the SCARED-41 for each informant.
6.
To examine criterion validity using the known-groups method, independent samples t-tests (for SCARED scores) and Fisher’s Exact Tests (for categorical demographics) were computed to compare participants classified as anxious versus non-anxious based on Anxiety Disorder status. Significant differences on SCARED scores in the expected direction would provide evidence for the scale’s ability to differentiate these clinically relevant groups.
7.
To evaluate the potential diagnostic accuracy of the SCARED-5 as a screening tool, receiver operating characteristic (ROC) curve analyses were conducted using the R pROC package. ROC analyses assess how well the SCARED-5 total score discriminates between anxious and non-anxious groups across its entire range. The Area Under the Curve (AUC) was calculated as a global measure of classification accuracy (interpreting >0.70 as fair, >0.80 as good, >0.90 as excellent (Swets et al., 2000). Anxiety group status (criterion) was defined first using the SCARED-41 total score cutoff and subsequently using the formal anxiety disorder diagnosis. Optimal cutoff scores on the SCARED-5 for potential screening purposes were determined using Youden’s J-statistic via the OptimalCutpoints package. This method identifies the score threshold that best balances sensitivity (the proportion of anxious youth correctly identified by the screen) and specificity (the proportion of non-anxious youth correctly identified by the screen). Classification statistics (sensitivity, specificity, positive/negative predictive power, overall accuracy) were computed for potential cut points (Kessel & Zimmerman, 1993; Youngstrom, 2014).
Results
Research site comparisons
Because approximately one-third of the sample was recruited from an anxiety clinic, we examined potential differences in demographic characteristics and anxiety scores across the three recruitment sites. While participant age did not differ significantly across sites (F(2, 106) = 1.92, p = 0.151), there was a significant difference in participant sex distribution (χ2(2) = 8.12, p = 0.017), with a higher proportion of males recruited from the allergy clinic (71.4%) compared to the general pediatric clinic (38.5%), with the anxiety clinic between but no different than the other two (51.4%).
As expected, one-way ANOVAs revealed significant differences across recruitment sites for mother-reported SCARED-5 scores (F(2, 105) = 19.65, p < 0.001), mother-reported SCARED-41 scores (F(2, 105) = 17.49, p < 0.001), child-reported SCARED-5 scores (F(2, 102) = 8.02, p < 0.001), and child-reported SCARED-41 scores (F(2, 102) = 10.33, p < 0.001). Tukey post-hoc tests for all four measures indicated that the anxiety clinic subsample had significantly higher mean scores than the allergy and the general pediatric clinics (all p < 0.003). For example, mean mother-reported SCARED-5 scores were 3.03 (SD = 2.26) for the anxiety clinic, 1.06 (SD = 1.30) for allergy, and 0.90 (SD = 0.99) for general pediatric clinics. Scores did not differ significantly between the allergy and general pediatric clinics on any measure (all p > 0.600).
Descriptive statistics
Item-level descriptive statistics for the SCARED-5 items are shown in Table 1. The rate of endorsing either 1 (somewhat true or sometimes true) or 2 (very true or often true) on the SCARED-5 items ranged from 22% to 56% for child ratings (M = 35.6%, SD = 14.0%) and from 16% to 40% (M = 25.4%, SD = 10.3%) for mother ratings.
Table 1
Study 1: SCARED-5 Item Endorsement and Factor Loadings
% Endorse
Factor
Item
0
1
2
Load
p
Child Self-Ratings
I get really frightened for no reason at all
76
17
7
0.44
0.001
I am afraid to be alone in the house
58
28
14
0.80
<0.001
People say I worry too much
65
19
15
0.33
0.003
I am scared to go to school
78
13
9
0.79
<0.001
I am shy
45
41
15
0.64
<0.001
Coefficient H reliability = 0.818
Omega reliability = 0.750
Mother Ratings of Child
Gets really frightened for no reason at all
84
13
3
0.67
<0.001
Afraid to be alone in the house
68
22
10
0.86
<0.001
People tell me that my child/adolescent worries too much
79
15
7
0.39
0.004
Scared to go to school
82
10
7
0.61
<0.001
My child/adolescent is shy
59
32
8
0.22
0.090
Coefficient H reliability = 0.813
Omega reliability = 0.697
0 = Not True or Hardly Ever True; 1 = Somewhat True or Sometimes True; 2 = Very True or Often True. Load = standardized factor loading. p = significance (p-value) of factor loading. Coefficient H reliability and Omega reliability were estimated using tools provided by McNeish (2018)
Confirmatory factor analyses
Confirmatory factor analyses of the SCARED-5 showed excellent fit for a one-factor model for child ratings, χ2(5) = 3.00, p = 0.6995, CFI = 1.00. RMSEA = 0.000, SRMR = 0.033, but not mother ratings, χ2(5) = 15.56, p = 0.0082, CFI = 0.844. RMSEA = 0.140, SRMR = 0.080. Modification indices showed model fit of mother ratings would improve by allowing correlated residuals between two items (“people tell me that my child/adolescent worries too much” and “my child/adolescent is shy”), suggesting these items share variance beyond what is accounted for by the overall anxiety factor, possibly reflecting a specific aspect of observable anxiety behavior. Re-estimating the model after this modification improved the model, resulting in an excellent fit, χ2(4) = 2.44, p = 0.6560, CFI = 1.00. RMSEA = 0.000, SRMR = 0.034. Standardized factor loadings and reliability estimates of factors are summarized in Table 1.
Correlations
Correlations showed that mother and child ratings of the SCARED-5 and SCARED-41 scores were significantly and highly associated within and between informants (see Table 2 inter-measure correlations). Regarding demographic correlates, neither participant age nor sex showed significant associations with mother or child reports on the SCARED-5 or SCARED-41 (all ∣r∣ ≤ 0.17, all p > 0.05).
Table 2
Study 1: Correlations Among SCARED-5 and SCARED-41 Scores from Mother and Child Reports
Variable
SCARED-5 Mother
SCARED-41 Mother
SCARED-5 Child
SCARED-41 Mother
0.86
–
–
SCARED-5 Child
0.42
0.51
–
SCARED-41 Child
0.43
0.56
0.84
All correlations are significantly different from zero at p < 0.05
Scale severity comparisons
Means scores on the SCARED-5 and SCARED-41 were compared using paired-sample t-tests and were significant for mother ratings, t(107) = −7.82, p < 0.001, Cohen’s d = −0.75, and child ratings, t(104) = −3.90, p < 0.001, Cohen’s d = −0.38. For both informants, anxiety scores were higher (more severe) for the SCARED-41 than the SCARED-5. Paired samples t-test also showed that mother and child ratings differed on the SCARED-5, t(103) = −3.61, p < 0.001, Cohen’s d = −0.35, and on the SCARED-41, t(103) = −3.42, p < 0.001, Cohen’s d = −0.34. Child ratings were higher (more severe) for both scores than mother ratings.
Comparisons of anxious and non-anxious youth
Youth with and without an anxiety disorder were compared on demographic measures and SCARED scores using independent samples t-tests for continuous measures and chi-square tests for categorical measures (see Table 3). Anxious and non-anxious groups did not differ on demographics. However, they differed in expected ways on both child and mother ratings on the SCARED-5 and SCARED-41, with effect sizes showing that the differences were large.
Table 3
Study 1 Comparison of Youth With and Without an Anxiety Disorder Diagnosis
No Anxiety
Yes Anxiety
Fisher’s Exact p
Effect Size
%
N
%
N
Male
54%
36 of 66
51%
22 of 43
0.845
Φ = −0.03
Hispanic
6%
4 of 66
12%
5 of 41
0.299
Φ = −0.11
Race
Asian
8%
5 of 66
2%
1 of 43
0.400
Φ = −0.11
Black/Af-Am
14%
9 of 66
5%
2 of 43
0.195
Φ = −0.15
Native Am
2%
1 of 66
0%
0 of 43
1.00
Φ = −0.08
White
86%
57 of 66
95%
41 of 43
0.195
Φ = 0.15
Other Race
0%
0 of 66
2%
1 of 43
0.394
Φ = 0.12
Multiple
8%
5 of 66
5%
2 of 43
0.701
Φ = −0.06
M (SD)
N
M (SD)
N
t-test
p
Effect Size
Age
12.01 (2.51)
66
12.81 (2.07)
43
t (107) = 1.73
0.086
d = 0.34
Mother ratings
SCARED-5
0.92 (1.13)
66
2.71 (2.20)
42
t (106) = 5.57
<0.001
d = 1.10
SCARED-41
13.50 (10.09)
66
27.81 (14.53)
42
t (106) = 6.04
<0.001
d = 1.19
Child ratings
SCARED-5
1.73 (1.98)
62
3.35 (2.35)
43
t (103) = 3.82
<0.001
d = 0.76
SCARED-41
18.72 (12.74)
62
30.30 (15.41)
43
t (103) = 4.20
<0.001
d = 0.83
Means (standard deviations in parentheses) are reported for continuous variables; percentages are shown for categorical variables. Sex: 0 = female, 1 = male. Hispanic: 0 = not Hispanic, 1 = Hispanic. Race: 0 = not in the racial group, 1 = yes in the racial group. Participants could select multiple race categories. Each category was compared independently between the No Anxiety and Yes Anxiety groups using Fisher’s Exact Test
Receiver operating characteristic analyses
ROC curves and AUC values (plus 95% confidence intervals) were computed for SCARED-5 scores in classifying youth as anxious or not, as defined by the SCARED-41 (Fig. 1) and as defined by having an anxiety disorder diagnosis (Fig. 2). For both outcomes, mother and child ratings were significant (AUC confidence intervals do not include 0.50), indicating that mother ratings and child ratings on the SCARED-5 accurately classified youth at greater-than-chance levels. In DeLong’s nonparametric test for correlated ROC curves, mother and child AUCs did not differ in magnitude (SCARED-41 p = 0.2066; anxiety disorder p = 0.5971).
Fig. 1
Study 1: Receiver Operating Characteristic Curve of SCARED-5 with SCARED-41 Group (No vs. Yes) as Criterion. Figure Notes: Sensitivity and 1-specificity are shown as percentages to aid in interpretation
Study 1: Receiver Operating Characteristic Curve of SCARED-5 with Anxiety Disorder Group (No vs. Yes) as Criterion. Figure Notes: Sensitivity and 1-specificity are shown as percentages to aid in interpretation
Youdan’s J-statistic indicated an optimal cutoff score of 3 for child SCARED-5 ratings and 2 for mother SCARED-5 ratings when SCARED-41 groups (anxious vs. non-anxious) were used as the criterion. The optimal cutoff score was 2 for both child and mother ratings when anxiety disorder group status (anxious vs. non-anxious) was the criterion. Classification statistics are summarized in Table 4 using both 2 and 3 as potential cut scores.
Table 4
Study 1: Diagnostic Classification Statistics for the SCARED-5
Cell Sizes
Chi-square
Classification Statistics
Criterion Groups:
Not Anx.
Anxious
SCARED-5 Groups:
Not Anx
Anx
Not Anx
Anx
Value
p
Sens
Spec
PPP
NPP
False Pos
False Neg
Overall Accuracy
SCARED-41 criterion
Mother
Cut score = 2
59
14
4
31
48.87
<0.001
88.6
80.8
68.9
93.7
19.2
11.4
83.3
Cut score = 3
67
6
11
24
42.95
<0.001
68.6
91.8
80.0
85.9
8.2
31.4
84.3
Child
Cut score = 2
45
17
2
41
47.39
<0.001
95.4
72.6
70.7
95.7
27.4
4.7
81.9
Cut score = 3
56
6
5
38
64.59
<0.001
88.4
90.3
86.4
91.8
9.7
11.6
89.5
Anx. dis as criterion
Mother
Cut score = 2
50
16
13
29
21.20
<0.001
69.0
75.8
64.4
79.4
24.2
30.9
73.2
Cut score = 3
58
8
20
22
20.74
<0.001
52.4
87.9
73.3
74.4
12.1
47.6
74.1
Child
Cut score = 2
39
23
8
35
20.15
<0.001
81.4
62.9
60.3
83.0
37.1
18.6
70.5
Cut score = 3
45
17
16
27
13.05
<0.001
62.8
72.6
61.4
73.4
27.4
37.2
68.6
Rows in italic font indicate the cutoff score supported by Youdan’s J-statistic
Anx Dis Anxiety disorder; Sens Sensitivity, the probability that SCARED-5 identifies anxiety given that the criterion identifies anxiety; Spec Specificity, the probability that the criterion does not identify anxiety given that criterion does not identify anxiety; PPP positive predictive power, probability of anxiety on the criterion given anxiety on the SCARED-5; NPP negative predictive power, probability of no anxiety on the criterion given no anxiety on the SCARED-5. False Pos percent of cases identified as anxious on the SCARED-5 but not anxious on the criterion; False Neg percent of cases identified as not anxious on the SCARED-5 but anxious on the criterion identifies anxiety; overall accuracy percent of cases identified as anxious on both measures plus the percentage of cases identified as non-anxious on both measures
Study 2
Study 2 was conducted to address two methodological limitations of Study 1. First, in Study 1, the SCARED-5 scores were derived from the full SCARED-41 administration rather than being independently administered. This approach mirrors the original method used by Birmaher et al. (1999), but it does not reflect how the scales would typically be used in clinical practice (as separate instruments) and may introduce context effects that influence item responses. Second, anxiety disorder classification in Study 1 relied partly on medical record review and the non-standardized ADQ. Study 2 addressed these limitations by: (1) administering the SCARED-5 and SCARED-41 as separate instruments several weeks apart, and (2) basing anxiety disorder diagnoses on comprehensive clinical evaluations of DSM criteria as determined by licensed psychologists using structured interviews and multiple information sources.
Method
Participants
Participants were 175 children (64.6% boys) ages 5 to 15 years (M = 7.97, SD = 2.33). Demographic information was collected via parent report. Parents reported their child’s race using a format requiring them to select the one category that best applied: one parent did not provide this information (N = 174 reporting). The reported racial distribution was as follows: Asian (n = 1, 0.6%), Black or African American (n = 4, 2.3%), Native American or Alaska Native (n = 1, 0.6%), White (n = 144, 82.8%), and Other or Multiple (n = 24, 13.8%). Regarding ethnicity, 12 participants (6.9%) identified as Hispanic or Latino. The median income for mothers was $40,000 to $50,000, and the median income for fathers was $60,000 to $70,000 (both ranged from $0 to over $100,000). Most participants had an attention-deficit/hyperactivity disorder diagnosis (ADHD; 91.4%), and a minority were diagnosed with an anxiety disorder (24.0%). Of those with an anxiety disorder, 37 had both anxiety and ADHD, while five had anxiety-only. Parents also completed the SCARED measures as part of the clinical assessment. For the SCARED-5, ratings were available from 173 mothers and 144 fathers (including 142 parent dyads who both provided ratings). For the SCARED-41, ratings were available from 171 mothers and 139 fathers (including 135 parent dyads).
Procedure
Data were collected as part of a comprehensive psychosocial evaluation conducted by a licensed Ph.D. psychologist. The Institutional Review Board approved using these clinical assessment data for research, waiving informed consent/assent because analyses were conducted retrospectively on existing clinical records. This study was not preregistered. As part of the evaluation, ratings were collected from the child and the child’s primary female and male caregivers (hereafter referred to as mother and father). Parents first completed the SCARED-41 part of the pre-visit intake paperwork and then filled out the brief SCARED-5 on the day of their child’s in-clinic assessment, on average 4–5 weeks later (Mother: M = 4.2, SD = 6.0; Father: M = 5.4, SD = 6.6). Children ages eight and above completed the SCARED-41 but not the SCARED-5; therefore, only mother and father ratings were used in this study. Psychiatric diagnoses were determined by a licensed clinical psychologist following criteria in the Diagnostic and Statistical Manual of Mental Disorders – Fifth Edition (American Psychiatric Association, 2013) using multiple sources of information, including rating scales completed by mothers, fathers, and teachers, a structured diagnostic interview with caregivers, a clinical interview with caregivers and children, and a review of school and health records. The same analyses were conducted as in Study 1.
Results
Assessment of Fixed Order Effects
Because all parents completed the SCARED-41 before the SCARED-5, we tested whether this fixed sequence biased the brief-form scores by regressing SCARED-5 scores on (a) baseline SCARED-41 scores and (b) the exact 4- vs 5-week interval between administrations. The interval coefficient serves as an order-effect probe -- if taking the long form first changes how parents respond to the brief form, a longer delay should systematically shift SCARED-5 scores after controlling SCARED-41 severity. The interval term was negligible and non-significant for mothers (β = –0.06, p = 0.31) and fathers (β = –0.06, p = 0.41). These findings suggest that the fixed administration sequence did not systematically bias parents’ responses on the brief measure, supporting the validity of comparing scores across the two forms despite their sequential administration.
Descriptive Statistics
Item-level descriptive statistics for the SCARED-5 are summarized in Table 5. Descriptive statistics for the SCARED-41 are in an online supplement (see Table S2). The rate of endorsing either 1 (somewhat true or sometimes true) or 2 (very true or often true) on the SCARED-5 items ranged from 15% to 44% (M = 28.4%, SD = 11.6%) for mother ratings and from 8% to 43% for father ratings (M = 26.6%, SD = 14.5%).
Table 5
Study 2: SCARED-5 Item Endorsement and Factor Loadings
% Endorse
Factor
Item
0
1
2
Load
p
Mother Ratings
Gets really frightened for no reason at all
72
18
10
0.91
<0.001
Afraid to be alone in the house
56
24
20
0.75
<0.001
People tell me that my child/adolescent worries too much
80
14
6
0.67
<0.001
Scared to go to school
84
13
2
0.70
<0.001
My child/adolescent is shy
65
25
10
0.38
<0.001
Coefficient H reliability = 0.890
Omega reliability = 0.823
Father Ratings
Gets really frightened for no reason at all
74
24
3
0.71
<0.001
Afraid to be alone in the house
57
26
17
0.67
<0.001
People tell me that my child/adolescent worries too much
83
15
2
0.82
<0.001
Scared to go to school
92
6
2
0.43
0.003
My child/adolescent is shy
62
31
7
0.39
<0.001
Coefficient H reliability = 0.811
Omega reliability = 0.750
0 = Not True or Hardly Ever True; 1 = Somewhat True or Sometimes True; 2 = Very True or Often True. Load = standardized factor loading. p = significance (p-value) of factor loading. Coefficient H reliability and Omega reliability were estimated using tools provided by McNeish (2018)
Confirmatory Factor Analyses
Confirmatory factor analyses of the SCARED-5 supported a one-factor model for mother ratings, χ2(5) = 3.03, p = 0.6958, CFI = 1.00. RMSEA = 0.000, SRMR = 0.026, and father ratings, χ2(5) = 10.24, p = 0.0687, CFI = 0.936. RMSEA = 0.080, SRMR = 0.078. Although fit statistics for the model using father ratings were modest, they support the one-factor model. Standardized factor loadings and reliability estimates of factors are summarized in Table 5. While both parent informants’ data supported a one-factor structure, the model fit was stronger for mother ratings, suggesting potential differences in how fathers perceive and report anxiety symptoms in their children.
Correlations
Correlations showed significant associations between SCARED-5 and SCARED-41 scores within and between mother and father informants (see Table 6 for inter-measure correlations). Examination of demographic correlates revealed no significant associations between participant age and any mother or father SCARED reports (all ∣r∣ ≤ 0.12, p > 0.05). For participant sex (coded 0=female, 1=male), only one significant association emerged: father report on the SCARED-5 was modestly positively correlated with male sex (r = 0.16, p < 0.05), indicating that fathers reported slightly higher scores for males compared to females on this version. Sex was not significantly correlated with mother reports on either scale or father reports on the SCARED-41 (other ∣r∣ ≤ 0.07, p > 0.05).
Table 6
Study 2: Correlations Among SCARED-5 and SCARED-41 Scores from Mother and Father Reports
Variable
SCARED-5 Mother
SCARED-41 Mother
SCARED-5 Father
SCARED-41 Mother
0.73
–
–
SCARED-5 Father
0.51
0.39
–
SCARED-41 Father
0.52
0.61
0.56
All correlations are significantly different from zero at p < 0.05
Scale Severity Comparisons
Paired-samples t-tests showed that SCARED-5 and SCARED-41 scores did not differ for mother ratings, t(170) = −0.48, p = 0.629, Cohen’s d = −0.04, or father ratings, t(132) = 0.35, p = 0.729, Cohen’s d = 0.03, indicating that ratings of anxiety were of equal severity when assessed by the SCARED-5 and SCARED-41. Paired sample t-tests comparing mother and father ratings did not differ for the SCARED-5, t(141) = 1.33, p = 0.187, Cohen’s d = 0.11, but did differ for the SCARED-41, t(134) = 2.69, p = 0.008, Cohen’s d = 0.23. On the SCARED-41, mothers rated anxiety as more severe than fathers, but the effect size shows this difference was small.
Comparisons of Anxious and Non-Anxious Youth
Youth with and without an anxiety disorder were compared on demographic measures and SCARED scores using independent samples t-tests for continuous measures and chi-square tests for categorical measures (see Table 7). Children in the anxious group were older, and more likely to be female and White. Mothers’ ratings on the SCARED-5 and SCARED-41 differed between anxious and non-anxious youth, with effect sizes indicating that the differences were medium to large. Notably, neither the SCARED-5 nor SCARED-41 completed by fathers significantly differentiated between youth with and without anxiety disorders, suggesting potential limitations in father-reported anxiety screening in this clinical population.
Table 7
Study 2: Comparison of Youth With and Without an Anxiety Disorder Diagnosis
No Anxiety
Yes Anxiety
Fisher’s Exact p
Effect Size
%
N
%
N
Male
71%
95 of 133
43%
18 of 42
0.001
Φ = −0.26
Hispanic
6%
8 of 133
10%
4 of 42
0.485
Φ = 0.06
Race
Asian
1%
1 of 132
0%
0 of 42
1.00
Φ = −0.04
Blk/Af-Am
3%
4 of 132
0%
0 of 42
0.573
Φ = −0.09
Native Am
1%
1 of 132
0%
0 of 42
1.00
Φ = −0.04
White
79%
104 of 132
95%
40 of 42
0.017
Φ = 0.19
Other/Multi
17%
22 of 132
5%
2 of 42
0.070
Φ = −0.15
M (SD)
N
M (SD)
N
t-test
p
Effect Size
Age
7.62 (2.30)
133
9.07 (2.11)
42
t (173) = −3.65
<0.001
d = 0.65
Mother ratings
SCARED-5
1.61 (1.99)
132
2.83 (2.50)
41
t (171) = −3.21
0.002
d = 0.57
SCARED-41
13.71 (11.38)
130
23.61 (16.02)
41
t (169) = −4.38
<0.001
d = 0.78
Father ratings
SCARED-5
1.66 (1.77)
110
1.56 (1.54)
34
t (142) = 0.31
0.755
d = −0.06
SCARED-41
11.90 (9.39)
109
15.20 (12.06)
30
t (137) = −1.60
0.112
d = 0.33
Means (standard deviations in parentheses) are reported for continuous variables; percentages are shown for categorical variables. Sex: 0 = female, 1 = male. Hispanic: 0 = not Hispanic, 1 = Hispanic. Race: 0 = not in the racial group, 1 = yes in the racial group
Receiver Operating Characteristic Analyses
A ROC analysis using the SCARED-5 to classify groups based on the SCARED-41 (see Fig. 3) showed that mother and father ratings on the SCARED-5 classified children at a level significantly above chance (AUC confidence intervals do not include 0.50). DeLong’s nonparametric tests showed that mother ratings had a significantly higher AUC than father ratings (p = 0.0458). When using anxiety disorder as the grouping criterion (see Fig. 4), mother ratings on the SCARED-5 classified children at a level significantly above chance, but father ratings did not (AUC = 0.49). This father rating AUC value of 0.49 is essentially at chance level, indicating that father-reported SCARED-5 scores provided no better discrimination between anxious and non-anxious youth than random guessing in this predominantly ADHD sample. Mother ratings had a significantly higher AUC than father ratings (p = 0.0089), suggesting substantial differences in the screening utility of the SCARED-5 depending on which parent completes the measure.
Fig. 3
Study 2: Receiver Operating Characteristic Curve of SCARED-5 with SCARED-41 Group (No vs. Yes) as Criterion. Figure Notes: Sensitivity and 1-specificity are shown as percentages to aid in interpretation
Study 2: Receiver Operating Characteristic Curve of SCARED-5 with Anxiety Disorder Group (No vs. Yes) as Criterion. Figure Notes: Sensitivity and 1-specificity are shown as percentages to aid in interpretation
Youdan’s J-statistic indicated an optimal cutoff score of 2 for SCARED-5 father ratings and 3 for SCARED-5 mother ratings when SCARED-41 group status (anxious vs. non-anxious) was used as the criterion. The pattern was reversed when anxiety disorder group status was the criterion; the optimal cutoff score was 3 for father ratings and 2 for mother ratings. Classification statistics are summarized in Table 8 using both 2 and 3 as potential cut scores.
Table 8
Study 2: Diagnostic Classification Statistics for the SCARED-5 using the SCARED-41 and Anxiety Disorder as the Grouping Criteria
Cell Sizes
Chi-square
Classification Statistics
Criterion Groups:
Not Anx.
Anxious
SCARED-5 Groups:
Not Anx
Anx
Not Anx
Anx
Value
p
Sens
Spec
PPP
NPP
False Pos
False Neg
Overall Accuracy
SCARED-41 Groups
Mother Ratings
Cut score = 2
97
35
2
37
57.71
<0.001
94.9
73.5
51.4
98.0
26.5
5.1
78.4
Cut score = 3
114
18
6
33
72.04
<0.001
84.6
86.4
64.7
95.0
13.6
15.4
86.0
Father Ratings
Cut score = 2
76
40
2
15
17.66
<0.001
88.2
65.5
27.3
97.4
34.5
11.8
68.4
Cut score = 3
95
21
6
11
17.62
<0.001
64.7
81.9
34.4
94.1
18.1
35.3
79.7
Anx. Dis. Groups
Mother Ratings
Cut score = 2
84
48
16
25
7.77
0.005
61.0
63.6
34.2
84.0
36.4
39.0
63.0
Cut score = 3
100
32
21
20
8.96
0.003
48.8
75.8
38.5
82.6
24.2
51.2
69.4
Father Ratings
Cut score = 2
63
47
20
14
0.03
0.873
41.2
57.3
23.0
75.9
42.7
58.8
53.5
Cut score = 3
82
25
28
9
0.01
0.906
26.5
74.5
24.3
76.6
25.5
73.5
63.2
Rows in italic font indicate the cutoff score supported by Youdan’s J-statistic
Sens Sensitivity, the probability that SCARED-5 identifies anxiety given that the criterion identifies anxiety; Spec Specificity, the probability that the criterion does not identify anxiety given that criterion does not identify anxiety; PPP positive predictive power, probability of anxiety on the criterion given anxiety on the SCARED-5; NPP negative predictive power, probability of no anxiety on the criterion given no anxiety on the SCARED-5. False Pos percent of cases identified as anxious on the SCARED-5 but not anxious on the criterion; False Neg percent of cases identified as not anxious on the SCARED-5 but anxious on the criterion identifies anxiety; overall accuracy percent of cases identified as anxious on both measures plus the percentage of cases identified as non-anxious on both measures
General Discussion
This two-study investigation provides the first comprehensive psychometric evaluation of the SCARED-5 as a brief anxiety screening tool for youth. Our findings support the unidimensional structure, internal consistency, and convergent validity of the SCARED-5 across informants. Mother and child ratings demonstrated good discriminative validity in distinguishing anxious from non-anxious youth, while father ratings showed limited utility. A cutoff score of 2 on the SCARED-5 completed by mothers or children appears to optimize screening sensitivity in clinical and community samples. These findings substantiate the SCARED-5 as an efficient screening tool while highlighting important informant differences. Table 9 summarizes the key psychometric findings across studies and informants.
Table 9
Summary of SCARED-5 Psychometric Properties Across Studies and Informants
Property
Study 1
Study 2
Child
Mother
Mother
Father
Structure & Reliability
CFA Model Fit
Excellent
Gooda
Excellent
Adequate
Factor Loadings Range
0.33–0.80
0.22–0.86
0.38–0.91
0.39–0.82
Coefficient H
0.818
0.813
0.890
0.811
Omega
0.750
0.697
0.823
0.750
Convergent Validity
Correlation with SCARED-41
0.84
0.86
0.73
0.56
Correlation with Other Informants
0.42 (M)
0.42 (C)
0.51 (F)
0.51 (M)
Mean Score Comparisons
SCARED-5 vs. SCARED-41b
Lower**
Lower**
No diff.
No diff.
Comparison Between Informants
Higher than M
Lower than C
No diff. from F
No diff. from M
Group Discrimination
Anxious vs. Non-anxious
Yes***
Yes***
Yes***
No
Effect Size (Cohen’s d)
0.76
1.10
0.57
-0.06
Diagnostic Accuracy
AUC (Anxiety Disorder)
0.76
0.76
0.72
0.49
Optimal Cutoff
2
2
2
3c
Sensitivity at Cutoff
81.4%
69.0%
61.0%
26.5%
Specificity at Cutoff
62.9%
75.8%
63.6%
74.5%
M Mother report; C Child report; F Father report AUC Area Under the Curve in Receiver Operating Characteristic analysis
**p < 0.001 ***p < 0.01
aGood fit after allowing correlated residuals between two items
bComparison of mean item scores (total score divided by number of items)
cNot recommended for clinical use due to chance-level classification accuracy
The SCARED-5 is a five-item rating scale developed to screen for anxiety in youth (Birmaher et al., 1999) Despite being used in many research studies and clinical settings since its introduction in 1999 (e.g., Chavira et al., 2004; Gjerde et al., 2023; Stepp et al., 2010; Waschbusch et al., 2020), little research has examined its psychometric properties. This validation study aimed to fill this gap and is particularly timely given recent recommendations from the U.S. Preventive Services Task Force supporting anxiety screening in youth (U. S. Preventive Services Task Force et al., 2022; Viswanathan et al., 2022).
Factor Structure and Internal Consistency
As summarized in Table 9, confirmatory factor analyses of the SCARED-5 showed that a one-factor model was strongly supported for child self-ratings and mother ratings and moderately supported for father ratings. The more modest support for the one-factor model in father ratings suggests potential differences in how fathers conceptualize anxiety symptoms in their children, which may contribute to the differential screening utility observed. The internal consistency reliability estimates were high across informants, supporting the computation of a single score from the SCARED-5 as a measure of anxiety.
Convergent Validity and Scale Severity
The SCARED-5 was significantly correlated with the SCARED-41 across all informants, supporting its convergent validity. The contrasting findings regarding scale severity between Studies 1 and 2 highlight how methodological differences can influence measurement outcomes. Study 1’s within-session administration showed higher severity on the SCARED-41 compared to the derived SCARED-5, while Study 2’s separate administration found comparable severity levels between independently administered scales. Several factors could explain this discrepancy. First, context effects in Study 1 may have influenced item interpretation - when embedded within the longer measure, the five SCARED-5 items might be interpreted differently than when presented alone. Second, item-order effects could play a role, as the position of the five SCARED-5 items within the full 41-item scale might affect responses. Third, the difference could reflect statistical regression to the mean, where extreme scores at first measurement tend to move closer to the average upon repeated measurement. This could explain the findings of Study 2, where the brief scale was administered weeks after the full scale. Fourth, the samples differed somewhat between studies (Study 1 included an anxiety clinic subsample while Study 2 was primarily an ADHD sample), which could contribute to different response patterns. Finally, response bias may differ between formats - respondents might approach a brief, focused screening tool differently than a comprehensive measure. Further research using counterbalanced administration order and comparable samples is needed to clarify severity equivalence across formats and informants. Based on these findings, the relationship between SCARED-5 and SCARED-41 scores appears context-dependent. When administered independently in Study 2, parent reports showed no significant differences in severity. However, this finding cannot be generalized to child self-reports or situations where the SCARED-5 items are embedded within the full measure.
Informant Differences
This study was the first to compare mother, father, and child ratings on the SCARED-5, and it provided important insights into differences between these informants. Mothers rated anxiety as less severe than that of youth but more severe than that of fathers. Mother-child correlations ranged from 0.42 to 0.59, while mother-father correlations ranged from 0.51 to 0.61, both consistent with previous research using the SCARED-41 (Behrens et al., 2019; Birmaher et al., 1999; Birmaher et al., 1997; Cosi et al., 2010; Dirks et al., 2014; Jansen et al., 2017; Linyan et al., 2008; Rappaport et al., 2017; Weitkamp et al., 2010; Wren et al., 2007; Wren et al., 2004).
Notably, when comparing anxious and non-anxious groups, both child and mother ratings significantly differentiated between groups with moderate to large effects (d = 0.57–1.10), while father ratings did not discriminate between diagnostic groups on either the SCARED-5 (d = −0.06) or SCARED-41 (d = 0.33). This pattern was further confirmed in ROC analyses, where father-reported SCARED-5 scores performed at chance levels (AUC = 0.49) in identifying youth with anxiety disorders, in stark contrast to mother ratings (AUC = 0.72). These consistent findings suggest that clinical decisions based solely on father-reported SCARED-5 scores should be made cautiously, particularly in populations with high rates of comorbid ADHD.
These informant differences may reflect varying exposure to the child’s anxiety symptoms, different thresholds for recognizing anxiety behaviors, or gender-related differences in attention to emotional symptoms. The reduced discriminative ability of father ratings is consistent with a previous study that used an extended version of the SCARED (Jansen et al., 2017) and suggests that fathers are less sensitive to differences between anxious and non-anxious groups. This may be especially pronounced for brief measures like the SCARED-5, highlighting that while short forms reduce respondent burden, this benefit may come at the cost of reduced clinical utility for certain informants or constructs (Ziegler et al., 2014).
Diagnostic Accuracy and Clinical Cutoffs
The ROC analyses provided valuable information about the SCARED-5’s classification accuracy. Both child and mother SCARED-5 ratings significantly discriminated between anxious and non-anxious youth, whether defined by SCARED-41 scores or clinical diagnosis, with AUC values (0.72–0.76) similar to those reported for other screening tools (Canals et al., 2012; Vigil-Colet et al., 2009) and for the SCARED-41 (Birmaher et al., 1997; Canals et al., 2012; Desousa et al., 2013; Hariz et al., 2013; Monga et al., 2000; Rappaport et al., 2017; Van Meter et al., 2018). This is noteworthy because the discriminative ability was supported when the SCARED-5 was derived from the SCARED-41 and when independently administered. Based on these analyses, a cutoff score of 2 optimizes sensitivity while maintaining specificity.
Clinical Implications
The findings from this study have several direct implications for clinical practice. First, when using the SCARED-5 as a screening tool in general pediatric or primary care settings, a cutoff score of 2 on either mother or child reports provides optimal sensitivity for identifying youth who may warrant further assessment. This is lower than the initially proposed cutoff of 3 (Birmaher et al., 1999), and reflects our prioritization of minimizing false negatives in screening contexts.
Second, clinicians should be aware of crucial informant differences when using the SCARED-5. Mother and child reports demonstrated good discriminative validity and should be prioritized when possible. In contrast, as noted in the Informant Differences section, father-reported SCARED-5 scores showed poor discriminative ability in our sample, suggesting they should be interpreted with considerable caution, particularly in populations with high rates of ADHD comorbidity. When only father reports are available, clinicians may need to supplement these with additional assessment methods or consider lowering the threshold for follow-up evaluation.
Third, the SCARED-5’s brevity makes it particularly valuable for time-constrained settings like primary care visits or school-based screenings where the full SCARED-41 may be impractical. Its strong correlation with the full measure suggests it can serve as an efficient first-step screening tool in a stepped assessment approach. However, clinicians should recognize that the SCARED-5 is not a diagnostic instrument; positive screens should trigger more comprehensive assessments that incorporate clinical interviews and consider differential diagnoses, comorbidities, and functional impairment.
Broader Implications and Applications
The validation of the SCARED-5 has implications that extend beyond its psychometric properties. In healthcare systems increasingly focused on integrated care models, a brief, validated anxiety screening tool addresses a critical need in time-constrained primary care settings. The SCARED-5’s brevity makes it feasible for implementation during brief pediatric visits, aligning with recent healthcare policy shifts that emphasize mental health integration and early identification (U. S. Preventive Services Task Force et al., 2022; Walkup et al., 2022). This is particularly relevant as primary care providers are increasingly expected to identify and address mental health concerns during standard well-child visits. Implementing the SCARED-5 in electronic health record systems could facilitate routine screening during these visits, potentially improving early identification of anxiety symptoms before they progress to more severe impairment.
The SCARED-5 also has potential applications in school-based mental health initiatives, where resources are often limited. Schools function as de facto mental health systems for many children (Santor et al., 2007; Waschbusch et al., 2005), and a validated 5-item measure could support efficient screening programs to identify at-risk students before symptoms progress to clinical levels. Furthermore, the measure’s brevity makes it well-suited for telehealth applications and digital platforms, which have expanded significantly in recent years and require efficient assessment tools (McGrath et al., 2011).
From an economic perspective, undetected anxiety disorders often lead to increased healthcare utilization, including emergency department visits for somatic complaints and school absenteeism (Ramsawh et al., 2010). The SCARED-5 could save healthcare costs through earlier identification and appropriate intervention. Additionally, the measure’s brevity makes it practical for repeated administration to monitor treatment progress or developmental trajectories, supporting measurement-based care approaches where interventions are adjusted based on systematic outcome monitoring (Waschbusch et al., 2020).
Our findings regarding informant differences also have important training implications for healthcare providers. Provider education should emphasize collecting maternal or child reports when possible, or supplementing father reports with additional assessment methods when screening for anxiety disorders. This nuance in informant utility should inform how practitioners implement and interpret anxiety screening in clinical settings, particularly for youth with comorbid conditions like ADHD.
Limitations and Future Directions
Several limitations should be considered when interpreting this study’s findings. Both samples had demographic and clinical characteristics that may limit generalizability. The samples were predominantly White (approximately 83% to 90%), limiting applicability to youth from other racial and ethnic backgrounds. Additionally, our samples had specific clinical compositions that may have influenced results. Study 1 included participants from three different recruitment sites, with approximately one-third from a specialized anxiety clinic who showed significantly higher anxiety scores than those from general pediatric and allergy clinics. Study 2 was primarily composed of youth with ADHD diagnoses (91.4%), with anxiety disorders present in only 24% of participants. This high rate of ADHD comorbidity might have influenced the psychometric performance of the SCARED-5, particularly for father reports, as ADHD symptoms can sometimes mask, mimic, or complicate the presentation of anxiety. This could potentially make it more difficult for certain informants to identify anxiety symptoms accurately. Future research should examine whether the SCARED-5’s psychometric properties, especially its informant patterns, are consistent across more diverse racial/ethnic populations and clinical compositions. Second, data regarding participation refusal rates were not systematically collected, limiting our ability to assess potential selection biases. Similarly, in Study 1, anxiety disorder diagnoses were partly based on the Anxiety Disorder Questionnaire, an investigator-designed measure without established psychometric properties. However, this limitation was addressed by corroborating diagnoses using medical records. Third, in Study 2, the fixed administration order (SCARED-41 first, then SCARED-5 several weeks later) was a methodological constraint. While our analyses suggested this sequencing did not systematically bias scores, future studies with counterbalanced administration would provide stronger evidence regarding potential order effects. Fourth, this initial validation study did not conduct advanced psychometric analyses such as Item Response Theory or Differential Item Functioning. Such analyses could provide valuable insights into whether the SCARED-5 functions equivalently across different demographic groups and whether the current five items represent the optimal selection from the SCARED-41 item pool.
Future research should examine the SCARED-5 in larger, representative community or general pediatric samples that better reflect the contexts where it would typically be used as a screening tool. Advanced psychometric analyses, such as Item Response Theory and Differential Item Functioning, should be applied to determine whether the current five items represent the optimal selection from the SCARED-41 item pool, and to evaluate measurement invariance across diverse demographic groups. Additional studies are needed to develop a comprehensive yet efficient process for screening anxiety in youth, potentially integrating the SCARED-5 into multi-stage assessment protocols. The cutoff recommendation should be considered preliminary until replicated in larger, more diverse samples.
In conclusion, this study provides strong evidence supporting the SCARED-5 as a psychometrically sound, efficient screening tool for anxiety in youth when completed by mothers or children themselves (see Table 9). While more research is needed to address the limitations identified, the current findings support the clinical utility of this brief measure in time-constrained settings where early identification of anxiety symptoms is crucial. The SCARED-5 represents an important tool in the broader effort to improve mental health screening and early intervention for youth, particularly as healthcare systems increasingly prioritize integrated approaches to addressing childhood anxiety.
The study was approved by the Institutional Review Board. Informed consent was collected from participants in Study 1. Informed consent was waived for participants in Study 2 because data were existing clinical records.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Met BSL Psychologie Totaal blijf je als professional steeds op de hoogte van de nieuwste ontwikkelingen binnen jouw vak. Met het online abonnement heb je toegang tot een groot aantal boeken, protocollen, vaktijdschriften en e-learnings op het gebied van psychologie en psychiatrie. Zo kun je op je gemak en wanneer het jou het beste uitkomt verdiepen in jouw vakgebied.
