A limited amount of research has been conducted on the psychometric properties of commonly used measures of anxious or depressive symptomatology for use on the internet, although such measures are seeing increasing use in internet administration for both clinical and research uses. A plethora of advantages exist for the use of internet administration of questionnaires, both in terms of assessment and the potential use in treatment monitoring as well as for research. The aim of this study was to examine the psychometric properties of two common clinical measures, the Penn State Worry Questionnaire and the. Depression, Anxiety, and Stress Scale, in an internet administered format (N = 1138). Results suggest that these two measures may be used with confidence in an online format in terms of reliability and validity.
The development and acceptance of internet technology has coincided with the emergence of internet-based questionnaire administration as a means of gathering psychological information. Community access to the internet has offered new opportunities in terms of assessing and treating mental disorders (Kobak, Reynolds, Rosenfeld, & Greist, 1990) and has opened a new frontier for research and data collection (Coles, Cook, & Blake, 2007). The United States Department of Commerce (2002) reported that 54% of all US citizens have access to the internet from their home. Further, research has found that one quarter of Americans with home internet access have used the internet for information on mental health conditions (Pew Internet, 2003). There are many obvious advantages to internet administration of psychological instruments including ease and comfort of administration, simplicity of data collection, computer facilitated scoring, as well as reduced cost and increased opportunity for researchers and clinicians to access larger and potentially more diverse populations (Coles et al., 2007, Naglieri et al., 2006, Reips, 2000 and Schmidt, 1997). Other advantages to web-based data collection include lower social anxiety or socially desirable responding demonstrated by participants (Joinson, 1999). Naglieri et al. (2006) suggest that although the use of the internet for psychological testing and assessment presents a number of advantages and benefits to both clinicians and researchers, it is not without its potential risks. For example, the implications a test has for an individual test taker may influence his/response style in a manner similar to that of paper-and-pencil measures.
Research on internet administration of self-report inventories suggest that reliable and valid data are gathered (Buchanan, 2000) and that psychometric properties of traditional measures are typically replicated although not identical (Buchanan, 2000) to paper-and-pencil formats. When data has been available for paper-and-pencil versions of online instruments, research has shown that the online measures do assess the same constructs as the traditional paper versions (Davis, 1999). In cases of instruments which have only an internet version available, there is evidence that the online measures possess construct validity (Buchanan, 2000 and Buchanan and Smith, 1999) in that they measure the traits proposed. Although there is evidence to suggest that online measures of psychological traits are valid and reliable, other research does suggest that the psychometric properties of the measure may change when administered via the web (Buchanan, 2000 and Buchanan, 2001). Naglieri et al. (2006) suggest that internet administration of psychological tests and measures lends itself to the use of Item Response Theory (IRT; Hambleton & Swaminathan, 1985). Interactive and fluid assessments through the internet can be tailored to the individual test-takers responses providing efficient and appropriate assessment. For example, an individual completing a measure of mental health may answer positively to a question querying depressive symptoms and additional questions screening for significant symptoms of depression will subsequently be presented whereas the individual denies a prominent symptom of psychosis and further questions on psychotic disorders are not presented. Overall, the research on internet administered self-report questionnaires suggest that the measures may be reliable and valid, but psychometric properties of web-based versions should not necessarily be assumed to be equivalent to paper-and-pencil counterparts.
The aim of this study was to examine the preliminary psychometric properties and normative data for the internet administered versions of two commonly used measures of depression and anxiety, the Penn State Worry Questionnaire (PSWQ) and the Depression, Anxiety, and Stress Scale (DASS). It was hypothesized that the psychometric properties of the measures (i.e. means, standard deviations, and reliability) would be similar to published reports utilizing traditional paper-and-pencil versions of the measures.
