Page 182 - Contributed Paper Session (CPS) - Volume 7

CPS2056 Nurul Hafizah Azizan et al.
dichotomous scale, category scale, numerical scale, itemized rating scale etc.
(Sekaran & Bougie, 2016). Hence, researchers need to carefully choose which
rating scale (i.e., types of rating scale, labels of rating scale and number scale
point) is suitable that can fulfil the psychometric properties and maximize both
validity and reliability of the survey instrument. Studies conducted by several
researchers have found that the characteristics of rating scale used such as
types of rating scale, labels of rating scale and number of response categories
with either odd or even response alternative might influence the quality of
measurement instrument (Allahyari, Jafari, & Bagheri, 2016; Cicchetti,
Showalter, & Tyrer, 1985; Daher, Ahmad, Winn, & Selamat, 2015; Eutsler & Lang,
2015; Osteras et al., 2008; Preston & Colman, 2000; Revilla, Saris, & Krosnick,
2014).
Gaskell, Wright and O’Muircheartaigh (1995) also remarked that the way
in which the respondents respond to the questions may substantially be
influenced by the construction of the response scale. In another study by
Preston and Colman (2000), they also claimed that reliability, validity and
discriminating power can be affected by the format of rating scale used. The
findings from this study suggested that with four‐point and a smaller number
of categories, the reliability, validity and discriminating power performed were
relatively poor but significantly higher with more response categories of up to
seven‐point. This reveals that besides playing a significant role on the quality
of response given by the respondents, the choice of suitable rating scales used
in the survey can also give a huge impact on the quality of the measurement
instrument in overall. However, Maydeu‐Olivares et al. (2009) said that the use
of optimal number of response alternatives for a survey is still a debate, and
no consensus about this issue has been reached. Apart from that, the use of
continuous rating scales has also received considerable attention by the
researches. One of the mostly used continuous rating scales is Visual Analogue
Scale (VAS). The use of a continuous rating scale in a survey has offered some
advantages. Since a wide range of scales are being provided, hence the
respondents will have the freedom to rate their response. Besides that, the
actual experience of the respondents can also be assessed (Hasson & Arnetz,
2005). Thus, the quality of the information gained from the survey could be
preserved.

3.1 Number of Response Categories and Its Effect on Validity and
Reliability
Numerous studies in the last few decades confirmed that by varying the
number of response categories may influence the validity and reliability of a
survey instrument. Osteras et al. (2008) conducted a study with randomized
design to compare the quality of instrument in terms of internal consistency
and discriminant validity with original 4‐point scale and new 5‐point scale

169 | I S I W S C 2 0 1 9