Elsevier

Journal of Affective Disorders

Volume 253, 15 June 2019, Pages 394-401
Journal of Affective Disorders

Research paper
Measurement invariance and psychometric analysis of Kutcher Adolescent Depression Scale across gender and marital status

https://doi.org/10.1016/j.jad.2019.05.010Get rights and content

Highlights

  • The Kutcher Adolescent Depression Scale (KADS) is a widely used self-report tool to diagnose and monitor clinical depression in adolescents.

  • All the items and the scale showed acceptable goodness of fit to the partial credit model (PCM).

  • Assessing measurement invariance across gender and marital status showed the adequacy of the KADS for measuring depression with practically negligible differences.

  • Results of this study provided further evidence of the utility and extendibility of the KADS to other population of participants.

Abstract

Background

The Kutcher Adolescent Depression Scale (KADS) has been studied across Canada and some other countries during the past almost 15 years. The scale is a self-report tool to diagnose and monitor clinical depression in adolescents. A brief review of previous studies on KADS showed the lack of evaluation of KADS fairness/equivalence in measuring depression among identified groups.

Methods

To examine the psychometric properties and measurement invariance of the KADS, responses of 407 participants were analyzed using item response theory (IRT) and ordinal logistic regression (OLR). Relevant measures of effect size were utilized to interpret the results.

Results

Findings of the parallel factor analysis confirmed unidimensionality of the KADS and the partial credit IRT model found to be the best fitting model for analyzing the scale. OLR analysis detected three items across gender and one item across marital status to function differentially. An assessment of effect sizes implied negligible differences for practical considerations.

Limitations

A note of caution is necessary with respect to interpreting results of measurement invariance across Gender. The sample analyzed in this study was predominantly female and this might have affected our findings. A similar analysis with a more balanced sample is recommended.

Conclusions

This study was a significant step towards providing theoretical and practical information regarding the assessment of depression among adolescents by presenting adequate evidence regarding the psychometric properties of KADS-11. Future studies may look at different methods for assessing invariance and different groups for strengthening conclusions with respect to the KADS.

Introduction

Depression is one of the prevalent mental disorders across Canada (Centre for Addiction and Mental Health (CAMH) 2011, Canadian Mental Health Association (CMHA) 2013, Kimball et al., 2018, Mood Disorders Society of Canada 2018) and around the world (World Health Organization 2015, World Health Organization 2017). This disorder, which is characterized by several symptoms such as depressed mood or irritable, loss of interest, sense of hopelessness, lack of happiness and self-withdrawal, psychomotor retardation and other symptoms (Steptoe et al., 2007, Walkiewicz et al., 2012), was predicted to be the second most common mental health disorder in Canada by 2020 (WHO, 2002 cited in CAMH, 2011). It is also estimated that 70% of mental health problems have their onset during childhood and adolescence (CAMH, 2015) and young adults are at risk for developing depression and other types of mental health difficulties (Canadian Mental Health Association (CMHA) 2013, Ialomiteanu et al., 2014). Thus, most clinicians, practitioners and counsellors endeavor to screen and diagnose depression as early as possible by relying on reliable and valid tools for further therapeutic intervention.

A recent review of most frequently used depression scales by Shojaee et al. (2016) revealed that amongst various tools, the Kutcher Adolescent Depression Scale-11 Items (KADS-11) (Brooks and Kutcher, 2001, Shahidi and Shojaee, 2014) is one of the most appropriate scales possessing several vital characteristics for assessing disorders in youth such as: 1) ease of administration; 2) adequate number of items; 3) the ability to distinguish comorbid symptoms; 4) specificity and sensitivity; 5) purpose: screening or diagnosis; 6) ability to measure change over time (treatment sensitivity); 7) internal reliability and validity; 8) developmentally appropriate; and 9) self-report rather than clinician administered (Brooks et al., 2003, Gesinde and Sanu, 2014, Levine, 2013, Roberge et al., 2013, Trujols et al., 2013).

The scale is a self-report tool to diagnose and monitor clinical depression in adolescents. Pursuing, the popular diagnostic criteria for depression such as DSM [IV-R and V], KADS has been made through central attention to some principal features such as: (a) items address the core symptoms of adolescent depression, (b) symptoms are described using both standard and colloquial terminology, and (c) items measure the frequency of occurrence (in the case of 14 items) or the severity (in the case of the other two items) of the specified symptoms (LeBlanc et al., 2002, p. 115).

Developing the first 16-item version, the primary studies of KADS started in 2001 (Brooks and Kutcher, 2001), and 2002 (LeBlanc et al., 2002). This version was studied on the different samples of students (aged 12–22 years old) attending secondary schools and adolescents who were referred to healthcare clinics for the treatment of Major Depressive Disorder (MDD). During 2002 and 2004, several studies were conducted in Canada to measure KADS’ sensitivity to change and its psychometric properties resulted in two other versions including KADS-11, and KADS-6 (Brooks and Kutcher, 2001, Brooks et al., 2003, Brooks, 2004).

For example, through using Canadian samples, LeBlanc et al. (2002) compared the diagnostic validity of KADS-16 items and 6-items with the Beck Depression Inventory (BDI) and the Mini International Neuropsychiatric Interview (MINI) in 1712 adolescents to determine the benefits of KADS and its psychometric characteristics. Recognizing which KADS items best identified subjects experiencing a Major Depression episode (MDE), LeBlanc used receiver operating characteristic (ROC) curve and found that the diagnostic ability of the short version of KADS was as good as BDI and was better than the full-length (16-item) scale.

From the time KADS was recommended for use in health care settings, including the Guidelines for Adolescent Depression in Primary Care in U.S.A and Canada (Jensen et al., 2012), the scale caught the researchers’ and clinicians’ attention and was studied in different countries such as Malaysia, Portugal, Iran, Poland, Nigeria, Jamaica, and Barbados (Kutcher and Chehil, 2007, Ahmad et al., 2011, Mojs et al., 2012, Mojs et al., 2013, Shahidi and Shojaee, 2014, Gesinde and Sanu, 2014, Shojaee et al., 2016, Lowe et al., 2018).

Using Malaysian sample in 2011, Ahmad and his colleagues studied the validity of the mental health component of locally used Adolescent Health Screening Form, KADS-6 items, Beck Anxiety Inventory, and Tool for Assessment of Suicide Risk: Adolescent Version (TASRS-A) to screen depression, anxiety and suicidal tendency. Results indicated that Adolescent Health Screening Tool was very sensitive but had low Positive Predictive Value (PPV) in the screening of anxiety symptoms. In detecting depressive symptoms using KADS 6-items, the findings showed high PPV but less sensitivity (i.e., 28.6%).

Mojs et al. (2012) focused on Polish samples and reported the internal consistency of KADS as measured by Cronbach's α at the value of 0.84. In 2013, Mojs and her colleagues conducted another research on the university students who were prone to depression using KADS and demonstrated that depression fluctuates in terms of the students’ major and course study, gender, and financial situation in which gender and financial situation were two influential variables in increasing the rate of depression in students.

Using Iranian samples in 2014, Shahidi and Shojaee conducted two studies, examining the psychometric properties of KADS 11-item version in one study and investigating the sensitivity of KADS-11 in another study. Both studies were conducted on an extended age, ranged from late adolescence age to young adults’ age, in order to measure whether the content of the scale is reliable for different ages. The idea was based on the new psycho-social perspectives demonstrating that adolescence period may fall between 10 and 26 years of age (American Psychological Association -APA 2002, Curtis, 2005, Meyer et al., 2008, Csikszentmihalyi, 2018). Some researchers in international organizations such as United Nation incorporated different age ranges (e.g., 10 to 25 by UNFPA or 15 to 32 by UN Habitat-Youth Fund) for defining adolescence and youth globally (United Nation-UNFPA 2017, United Nations Department of Economic and Social Affairs-UNDESA 2015). Although this extended age range is not fully in agreement with other documents (APA, 2002), Shahidi and Shojaee (2014) postulated that such psychological screening scales may be relevant and reliable not only for adolescents (aged from 11 to 18 years) but also for individuals in late adolescence and young adults (19–32). Through this postulation, the first study on 277 Iranian university students showed that the test has appropriate a reliability of 0.88 (i.e., Cronbach's α) for all items. In the second study, Shahidi and Shojaee (2014) administered the KADS on 63 individuals diagnosed as depressed and 40 non-depressed individuals to compare their responses. The analysis of the data revealed that the scale has enough power to differentiate depressed group from non-depressed group (i.e., t (101) = 3.316, p < 0.001). Similar results were found for both factors.

Increasing the popularity of KADS in these years, this scale (KADS-11) was also used with 550 Nigerian university students by Gesinde and Sanu (2014). Their analysis was based on the cutoff score of 11 for moderated and 23 for severe depression. They reported that the prevalence of depression was ranged from 11.45% to 35.81% for both genders (female and male) and there was no significant difference between male and female depression level. Using Item Response Theory (IRT) by employing multidimensional form of Graded Response Modeling (GRM), Shojaee et al. (2016) examined the relationship between item responses and the latent trait of the KADS-11. Analyzing the responses of 300 Iranian university students brought to light that almost all 11 items of KADS have the power to provide maximum amount of information regarding the scale's latent trait. In addition, it was found that 55.2% of the total variance of the KADS-11 can be explained by two extracted core factors, Core Depression factor and Suicidal and Physical factor. Finally, the psychometric properties of the scale were investigated, that is the reliability of the scale as a coefficient alpha reached at 0.88.

The most recent study on KADS was conducted on 759 Caribbean samples (Lowe et al., 2018). Using the version of 11-item of KADS in their study, Lowe et al. (2018) demonstrated that the scale has an acceptable degree of reliability (α = 0.76) and reasonable concurrent validity through high correlation with the Adolescent Depression Rating Scale (ADRS). The scale had also significant discriminant validity through low correlations with the state and trait anxiety visual analogue scales.

A brief review of these previous studies on KADS shows the lack of evaluation of KADS fairness/equivalence in measuring depression among identified groups. Assessing measurement equivalence also known as differential item functioning (DIF), is an important part of the process of validating questionnaires to test whether the probability of responding to a specific item exhibits different statistical properties for different identifiable groups after matching the groups on an ability measure (De Ayala et al., 2002, Mousavi and Krishnan, 2016, Sharafi et al., 2017). Therefore, the goal of this study was to assess measurement equivalence of depression by means of KADS across gender and marital status.

Section snippets

Sample

This study involved 407 students who were selected using a systematic multistage random sampling method from Islamic Azad University-Tehran Central in Iran. Of this group of participants, 45 (11.1%) individuals were males and 362 (88.9%) were females. Additionally, of the total participants were 301 (74%) single and 106 (26%) were married. The mean age of the sample was 22.6 years (SD = 3.0). All students were studying either psychology or social sciences.

Instrument

The Kutcher Adolescent Depression Scale

Results

Result of parallel factor analysis is illustrated in Fig. 1. It can be clearly seen that there is one dominant factor and several smaller factors. This finding suggested a one-dimensional structure for the KDAS. Furthermore, there were two factors with eigenvalues greater than 1 of which the first factor accounted for 46% and the second factor accounted for only 9% of the total variance.

Conclusion

Increasing evidence has shown the adverse relationship between depression and mental health outcomes and the necessity of using reliable and valid measurement tools (Hsu et al., 2014). Assessing improvements of psychiatric outcomes is especially difficult because mental disorders cannot be fully quantified using standard and laboratory procedures (Zimmerman et al., 2008). This statement further implies the requirement of detailed scrutiny of psychological tests particularly for specific target

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Conflict of interest

Author declare no conflict of interest.

CRediT authorship contribution statement

Amin Mousavi: Writing - original draft, Formal analysis. Mahnaz Shojaee: Methodology. Mehrdad Shahidi: Data curation, Project administration. Ying Cui: . Stan Kutcher: Writing - original draft.

Acknowledgements

None.

References (61)

  • S.J. Brooks et al.

    Diagnosis and measurement of adolescent depression: a review of commonly utilized instruments

    J. Child Adolesc. Psychopharmacol.

    (2001)
  • S.J. Brooks et al.

    The Kutcher Adolescent Depression Scale: assessment of its evaluative properties over the course of an 8-week pediatric pharmacotherapy trial

    J. Child Adolesc. Psychopharmacol.

    (2003)
  • Fast Facts about Mental Illness

    (2013)
  • Mental Health Facts and Statistics

    (2011)
  • Mental Health Facts and Statistics

    (2015)
  • R.P. Chalmers et al.

    It might not make a big DIF:improved differential test functioning statistics that account for sampling variability

    Educ. Psychol. Meas.

    (2016)
  • Csikszentmihalyi, M. (2018). Adolescents. Encyclopedia Britannica...
  • A.C. Curtis

    Defining adolescence

    J. Adolesc. Fam. Health

    (2005)
  • R.J. De Ayala et al.

    Differential item functioning: a mixture distribution conceptualization

    Int. J. Test.

    (2002)
  • R. Estabrook et al.

    Differential item functioning in the Cambridge Mental Disorders in the Elderly (CAMDEX) Depression Scale across middle age and late life

    Psychol. Assess.

    (2015)
  • W.H. Finch et al.

    Detection of crossing differential item functioning: a comparison of four methods

    Educ. Psychol. Meas.

    (2007)
  • A.M. Gesinde et al.

    Prevalence and gender difference in self-reported depressive symptomatology among Nigerian university students: implication for depression counselling

    Counsellor

    (2014)
  • J. Gómez-Benito et al.

    Efficacy of effect size measures in logistic regression: an application for detecting DIF

    Methodology

    (2009)
  • L.G. Humphreys et al.

    An investigation of the parallel analysis criterion for determining the number of common factors

    Multivar. Behav. Res.

    (1975)
  • A.R. Ialomiteanu et al.

    CAMH Monitore ReportSubstance Use, Mental Health and Well-Being among Ontario Adults, 1977–2013 (CAMH Research Document Series, 40)

    (2014)
  • Jensen, P.S., Cheung, A., Zuckerbrut, R., Ghalib, K., & Levitt, A. (2012). Guidelines for adolescent depression in...
  • M.G. Jodoin et al.

    Evaluating Type I error and power rates using an effect size measure with logistic regression procedure for DIF detection

    Appl. Meas. Educ.

    (2001)
  • R.N. Jones et al.

    Differential item functioning in patient reported outcomes measurement information system (PROMIS®) physical functioning short forms: analyses across ethnically diverse groups

    Psychol. Test Assess. Model.

    (2016)
  • S.M. Kimball et al.

    Database analysis of depression and anxiety in a community sample—response to a micronutrient intervention

    Nutrients

    (2018)
  • S. Kutcher et al.

    Gestão de risco de suicídio. Um manual para profi ssionais de saúde

    (2007)
  • View full text