Performance analysis and quality monitoring are essential to any organization to ensure programs and different organizational activities are all aligned and functioning according to the desired plan. One of the best indications and commonly used is patients’ satisfaction and feedback; hence, the widespread use of surveying techniques. Likert scales, especially, have been very helpful in this regard due to their simplicity, ease to conduct and interpret, and high specificity as to which aspects are being measured. In fact, surveying before and after implementing changes proves very helpful as it indicates the level of success or failure. So, what is the composition of a Likert scale and what are the opportunities and limitations of this survey instrument?
Origin and composition of Likert scale
The way attitudinal research has changed greatly since Likert, R. introduced his scale in 1932 (Edmondson, Edwards, & Boyer, 2012). It uses the usual bipolar response points that today almost everyone is familiar with. According to Edmondson, Edwards, and Boyer (2012), Likert scale has proved more reliable and used only half the number of items used by other popular methods at the time of its introduction. Also, in most cases, the scale does not require judges which added to its simplicity. Since then, Likert scale has become very popular and many modifications have been added to enhance the accuracy of the results (Edmondson, Edwards, & Boyer, 2012).
Likert scale is composed of the analysis of a series of statements to which respondents agree or disagree with variable degrees (Albaum, 1997). According to Edmondson, Edwards, and Boyer (2012), originally, Likert scale was composed of 4 different components. First one was strictly a yes/no/unsure question. The second is a simple multiple choice question. The third is composed of a statement followed by five response options of strongly agree/agree/undecided/disagree/strongly disagree. The fourth component is has the same options like the third one but the statement has a conflicting fact. Interestingly, Albaum (1997) described a different model of Likert scale composed by distinguishing between attitude direction and the strength of that attitude. This requires a first question to whether the respondent agree or disagree followed by a second question about the strength of his/her agreement or disagreement.
Development of questions
According to Edmondson, Edwards, and Boyer (2012), there are four steps to compose a likert scale: generating the items, assigning ratings to items, selecting the items, and administering the final scale. The rating step requires judges to give a specific value on a scale from 1 to 5 according the relevance of the item to subject researched. Also, noted by Edmondson, Edwards, and Boyer (2012), the number of scale points have been modified from the originally 5 introduced by Likert, R.
According to Grogan, Conner, Willits, and Norman (1995), in the UK, general practitioners have...