Assessing users satisfaction with web digital library: the case of Universiti Teknologi MARA

Abstract

Purpose – The purpose of this paper is to examine the determinants of user satisfaction in the context of academic web digital library (DL). A model based on the re-specified information system success model was developed and tested using the structural equation modeling (SEM) technique. Design/methodology/approach – The study employed survey research methodology with self-administered questionnaire as the research instrument. The questionnaire was developed based on the instruments used by previous researchers. The population of the study was students enrolled for the bachelor’s degree in the Faculty of Information Management, Universiti Teknologi MARA, Malaysia. These students were chosen because of researcher’s easy access to the sampling frame. Descriptive analysis and inferential analysis which include SEM were executed using IBM SPSS and AMOS statistical software. Findings – The findings indicate that information quality, systems quality, service quality, perceived usefulness, perceived ease of use and cognitive absorption are significant predictor of users’ satisfaction with the web DL. Research limitations/implications – Instead of collecting data from students from various faculties, this study only covered students enrolled in one faculty. In the same light, as the respondents were from one faculty, they only report their experience of using one DL. Given these limitations, the results obtained are narrowed in terms of generalizability. The second limitation of the study is related to the predictor variables which only focussed on six variables only. Practical implications – From the practical viewpoint, the instrument that has been developed can be used as a diagnostic tool for continuous improvements of the DL. Originality/value – From the theoretical viewpoint, the study has developed an empirical based framework that depicts critical factors influencing DL satisfaction. Researchers specializing on the assessment of DL effectiveness can consider adopting the framework for future studies. Alternatively, the framework can be further extended by integrating other variables such as users’ characteristics or organizational characteristics.

A Rapid Review of the Reporting and Characteristics of Instruments Measuring Satisfaction with Reference Service in Academic Libraries

“…Luo and Buer's instrument (2015) included 10 components to express satisfaction but did not address another potential component: ethical issues as identified by Kloda and Moore's (2016) question item, "The consult reflected a respect for my confidentiality as a library user." On their survey measuring satisfaction with digital library service, including virtual reference, Masrek and Gaskin (2016) included 24 items representing 6 component factors of satisfaction, in addition to three items related to overall satisfaction (Masrek & Gaskin, personal communication, April 20, 2017). Instrument developers might consider that responses to a single question about satisfaction are likely to be positive because "providing tailored individual help … will always be appreciated, which skews user satisfaction in survey results" (Fournier & Sikora, 2015, p. 255).…”

“…Blake et al (2016) provided evidence of face validity, item validity, and construct validity within their study; Huang et al (2015) demonstrated testing for face validity and construct validity; and Masrek and Gaskin (2016) also showed evidence of a combination of face validity and construct validity. Blake et al changed their survey "to reflect the responses received from librarian reviews and campus interview sessions," and consulted instrument development experts who helped them address content (item) and internal structure (construct) validity components (p. 227).…”

“…" Yan, et al (2015) were not clear, as it seemed they gave their scale two midpoint labels, "Neutral" and "Not Familiar." Most 5-point scales had a midpoint labeled "Neutral" (Askew, 2015;Blevins, et al, 2016;Boyce, 2015;Huang et al, 2015;Mohindra & Kumar, 2015) or "Neither Agree nor Disagree" (Jacoby et al, 2016;Masrek & Gaskin, 2016). Three authors did not report the label used for their midpoint (Chen, 2016;Ganaie, 2016;Swoger & Hoffman, 2015).…”

“…Most (83%) of the studies did not state if they had tested their instruments for reliability. Five articles (17%) (Chen, 2016;Dahan, et al, 2016;Huang et al, 2015;Masrek & Gaskin, 2016;Yan, et al, 2015) reported measurement of internal consistency for each component of the satisfaction construct when the component was measured by multiple scale items. All of the researchers used Cronbach's ɑ (alpha) test of internal consistency, in which a value of 0.70 is commonly believed to be a basic threshold of acceptable level (Nunnally, 1978).…”

Untitled

“…In libraries, regression [4] and structural equation modeling (SEM) are currently applied to estimate user satisfaction with RFID [13,14]. There are different SEM modeling estimation procedures, with the most frequently used being the maximum likelihood (ML) and partial least square (PLS) estimators.…”