Research report
The development and psychometric investigation of the university student depression inventory

https://doi.org/10.1016/j.jad.2006.05.007Get rights and content

Abstract

Background

Depression is prevalent among university students. It is debilitating and affects the academic performance of students. One of the most common reasons why students consult counselling centres is depression. Therefore it is important to assess the university students' depression by valid measures designed particularly for this purpose.

Method

The present study developed the “University Student Depression Inventory” (USDI) to measure the depressive symptoms of university students.

Result

Items based on a broad range of depression symptoms were subjected to factor analysis, resulting in a 30 item scale with three subscales: Lethargy; Cognitive/Emotional; and Academic Motivation. Psychometric investigation revealed satisfactory reliability and divergent and convergent validity.

Conclusion

It is expected that multidimensional USDI would be a useful tool for university counselling services to assess the depression of the students.

Introduction

Depression is a common emotional disorder, which causes distress and impairs functioning. The cognitive, physiological, behavioural and motivational symptoms are the core features of this disorder (Beck, 1967) and different combinations of these symptoms are experienced by depressed individuals (Hair et al., 1998). The severity of these symptoms of depression can range from mild to severe (Angst and Merikangas, 1997). While depression is a community wide problem which affects many people of all ages, gender and cultures (APA, 2000), university students (the term refers to university and college students) also commonly experience depression. Estimates of depression in the university population ranges from 30% of students experiencing some level of depression to around 15% of students experiencing clinical levels of depression at any one time (McLennan, 1992, Rosenthal and Schreiner, 2000).

Depression is a serious problem for university students. Depressed students are shown to have problems with university academic work and motivation, and report receiving lower grades than non-depressed students (Lyubomirsky et al., 2003, Vredenburg et al., 1988). Depression also causes more global problems for students, such as decreased quality of life and increased risk of suicide (Beck and Young, 1978, Simpson et al., 1996). Keeping in view the problematic nature of students' depression it is essential that university counselling centres can adequately assess depression in the student population.

While depression has been widely studied in clinical population, there has been limited research that specifically investigates this mood disorder in university students. The university environment presents many challenging and demanding situations, such as adjusting to a new environment, mastering new skills (Martin et al., 1999) and frequent academic stress (Misra and Castillo, 2004) that may elicit depression in students. Vredenburg et al. (1988) performed one of the few studies on student depression and found that even though the severity of students' symptoms was mild, their depression was chronic and caused ongoing impairment, including increased suicide risk (Beck and Young, 1978). Students' problematic symptoms included cognitive symptoms such as perfectionist ideation, worthlessness and low self-esteem, motivational symptoms such as dependency and loss of initiative, and social problems surrounding making friends and assertiveness (Vredenburg et al., 1988). These symptoms, which were also a common manifestation of subclinical depression (Beck, 1967, Rapaport et al., 2002) debilitated university students (Vredenburg et al., 1988).

Depression among university students, in spite of its alarming nature, has not been investigated extensively. A review of the literature indicates that students have been used as a convenience sample in analogue studies to examine clinical depression. Further, university students have been used in factor analysis studies to explore the factor structures of various clinical depression scales. Results from these studies indicate that although student depression is qualitatively similar to clinical depression, there may be slight differences in the manifestation and the severity of the symptoms. Subtle quantitative differences are primarily related to symptom severity (Cox et al., 1999, Cox et al., 2001). Researchers have also indicated that there are minor differences in the specific symptoms experienced by students. For example, students' depression is more cognitive than clinical depression (Cox et al., 1999, Whisman et al., 2000). Students, when depressed experienced lack of concentration, pessimism, self-blame, self-dislike and lack of energy (Cox et al., 1999). These studies also indicate that some major somatic symptoms of clinical depression, such as changes in sleep and appetite, are not useful as indicators of depression in students as they are caused by other factors such as social and academic schedules (Kitamura et al., 2004, Smith et al., 2001, Steer and Clark, 1997). Overall, students' depression is similar to clinical and sub-clinical depression, with some special features.

The major limitation of analogue and factor structure studies using student samples is that they are not designed to measure symptoms and severity of student depression in itself; rather they are designed to observe clinical symptoms in students with high ratings on clinical depression scales. As a consequence, these studies may only be investigating students with higher severity of clinical symptoms, rather than students who experience problematic depression.

Student depression is one of the most common reasons why students visit university counselling centres (Benton et al., 2003, Surtees et al., 1998). Up to 40% of students visiting university counselling centres have problems with depression, and the demand for depression services is increasing (Benton et al., 2003, Surtees et al., 1998). Moreover, many students seeking university counselling services for other issues are also experiencing depression (Surtees et al., 1998). In order to manage the high demands for services, university counselling centres must be able to identify and deal with depression when necessary.

While counsellors are able to detect depressive symptoms in students through clinical interview, it may be difficult if depressed mood is not the primary reason for the consultation or if counselling time is limited. Smith et al. (2001) suggest using a screening tool for depression, in particular to detect those students who come to university counselling services with depression as a hidden problem. They suggest that a screening tool is more useful than clinical interview, particularly if mood is not the primary symptom. The use of a scale also enables counsellors to quickly identify and address specific problems that the student is experiencing. Further, it can also be used to evaluate the effectiveness of an intervention.

Some university counselling centres have developed their own scales to measure depression in students (e.g. Benton et al., 2003). However, the drawback of using a clinic-developed scale is that they are not psychometrically assessed, and therefore the accuracy and clinical utility of the scale has not been objectively tested. The College Adjustment Scales (Anton and Reed, 1991) is a scale specifically designed to assess numerous problems that university students commonly experience, including depression and suicide. However, subsequent research on the scale has questioned its psychometric properties and its ability to distinguish between problem subcategories (Campbell and Pritchard, 2000). Alternatively, scales such as Depression scale of the Depression Anxiety Stress Scale (DASS) was partially developed on students to assess whether a person's distress was primarily caused by depression, anxiety or stress (Lovibond and Lovibond, 1995). Although it measures depression in general, it may not capture the specific cognitive, motivational and social symptoms experienced by depressed students.

Clinical depression scales such as the Beck Depression Inventory-II and Zung scale have excellent psychometric properties, however they are designed for use with clinical populations (Beck et al., 1996, Zung and Durham, 1965). As pointed out earlier, these scales are not necessarily indicative of level of depression in a student sample, and therefore may not identify students who are experiencing problematic depression. Secondly, these scales also include items related to appetite and sleep patterns. Fluctuations in these particular prominent symptoms of depression have been found in university students, regardless of level of depression (Smith et al., 2001, Steer and Clark, 1997). Further, a recent study examining the factor structure of Zung scale, using a large university student sample, has pointed out that “sleep disturbance” is not an indicator of students' depression (Kitamura et al., 2004). Additionally, clinical scales only measure major clinical symptoms, specifically emotional, cognitive and physiological depression symptoms. However, depressed students have been shown to have many problems with motivation, another aspect of depression (Beck, 1967, Vredenburg et al., 1988). Motivation is a major aspect of university life and identifying the nature of depressed students' motivational problems is important in assessment, because these symptoms would directly impact their success at the university. Furthermore, if university services employed clinical scales they would not pick up students with subclinical levels of depression.

Thus, university students frequently experience depression, which may range from mild to moderate in degree. Further, their manifestation may be characterised by specific cognitive, motivational and social symptoms relevant to the educational environment. In order to identify and manage such students, university counselling services need a psychometrically sound scale which adequately reflects students' depression. The aim of the current study was to develop the University Student Depression Inventory (USDI), a scale which adequately measures student depression and reflects the manifestation of student depression as cited in the literature. The second aim of the study was to assess the psychometric properties of the scale.

Section snippets

Participants

The sample consisted of 322 students who were studying at the Queensland University Technology in Brisbane, Australia. Participants included 247 (76.7%) women and 74 (23.0%) men, whose mean age was 24.44 years (SD = 7.79 years; 1 participant did not indicate gender; 4 participants did not indicate age). Participants were studying in the schools of psychology (63.0%); humanities and human services (20.2%); health (5.9%); business (3.7%) and others (6.8%). Students were predominantly from first

Data cleaning and screening

Less than 5% of the sample had missing data, and subjects with missing data on the USDI were deleted so as to avoid overfitting the data in factor analysis (Tabachnick and Fidell, 2001). Subjects with missing data on other scales were excluded from those analyses. Some items had skew and univariate and multivariate outliers, these items were transformed however the original data was used as transformation did not change the output.

Internal consistency of DASS and LSS–S

Coefficient alphas for the DASS total, Depression, Anxiety and

Discussion

The University Student Depression Inventory (USDI) was developed to measure depression among university students. To develop this scale, items were generated based on students' and university counsellors' experiences of student depression. Factor analysis resulted in a 30 item psychometrically sound inventory of student depression with three interrelated subscales: Cognitive/Emotional, Lethargy and Academic Motivation.

Psychometric properties

Psychometric analyses were performed on the USDI and its subscales to ensure the reliability and validity of the scales. To ensure that all sources of error for a self-report measure were estimated, both internal consistency and temporal stability were calculated (Henson, 2001). The two reliability measures were good for the full scale and all subscales. High test–retest correlations over 1 week reflected the stability of student depression, and demonstrated low error in the scale's measurement

Acknowledgements

The authors would like to thank the students, enrolled in the Psychological Assessment unit in 2003, for assisting in the generation of item pool and the QUT Student Counsellors for their helpful comments in process of short listing items.

References (40)

  • S.A. Benton et al.

    Changes in counseling center client problems across 13 years

    Prof. Psychol. Res. Pract.

    (2003)
  • M.H. Campbell et al.

    Factor structure of the College Adjustment Scales

    Psychol. Rep.

    (2000)
  • R.B. Cattell

    The Scree Test for the number of factors

    Multivariate Behav. Res.

    (1966)
  • R.F. DeVellis

    Scale Development: Theory and Applications

    (1991)
  • R.L. Gorsuch

    Factor Analysis

    (1983)
  • J.F. Hair et al.

    Multivariate Data Analysis

    (1998)
  • R.K. Henson

    Understanding internal consistency reliability estimates: a conceptual primer on coefficient alpha

    Meas. Eval. Couns. Dev.

    (2001)
  • H.F. Kaiser

    An index of factorial simplicity

    Psychometrika

    (1974)
  • T. Kitamura et al.

    Factor structure of the Zung Self-rating Depression Scale in first-year university students in Japan

    Psychiatry Res.

    (2004)
  • P. Kline

    A Handbook of Test Construction: Introduction to Psychometric Design

    (1986)
  • Cited by (38)

    • Measurement invariance of the Multidimensional Depression Assessment Scale (MDAS) across gender and ethnic groups of Asian, Caucasian, Black, and Hispanic

      2022, Journal of Affective Disorders
      Citation Excerpt :

      Centre for Epidemiologic Studies Depression Scale (CESD) contains two interpersonal items and could be insufficient in capturing detailed cultural variations. The well-validated University Student Depression Inventory (USDI) by Khawaja and Bryden (2006) also contained only two interpersonal symptoms. However, the Multidimensional Depression Assessment Scale (MDAS) fills the gap by entailing a comprehensive spectrum of interpersonal dysfunctional manifestations and motivations to measure depression of different severity.

    • Depressive symptoms in first year students at a rural South African University

      2020, Journal of Affective Disorders
      Citation Excerpt :

      Considering that the research was focused on adolescent students, the students in the class over the age of 21 years were excluded. The University Student Depression Inventory (USDI), which was used in this study, was developed by Khawaja and Bryden (2006) and has been successfully employed to date (Lee et al., 2013; Pillay and Bundhoo, 2015). The instrument is a 30-item self-report inventory that asks participants to respond to items assessing depressive symptoms, e.g. ‘I feel worthless’; ‘I have thought about killing myself’; ‘I think most people are better than me’.

    • Depression among Indian university students and its association with perceived university academic environment, living arrangements and personal issues

      2016, Asian Journal of Psychiatry
      Citation Excerpt :

      Based on their feedback the items were refined and the questionnaire was finalized. University Student Depression Inventory (USDI) (Khawaja and Bryden, 2006): This 30-item scale, with three factors, measures depressive symptoms among university students. Depression is manifested by cognitive, emotional, lethargy, and academic motivational symptoms.

    • University student depression inventory (USDI): Confirmatory factor analysis and review of psychometric properties

      2013, Journal of Affective Disorders
      Citation Excerpt :

      The analysis of the psychometric properties also established the USDI as a reliable and valid scale. This is in line with previous studies (Habibi et al., 2011; Khawaja and Duncanson, 2008; Khawaja and Bryden, 2006; Santos et al., 2008). The internal consistency of the total scale as well as the three factors was good, with high Cronbach alpha values.

    View all citing articles on Scopus
    View full text