For the construction of measurement instruments, rating scales are very important whether one talks of applied research or basic business research. There is a lot of regular and crucial dependence on rating scales and for this reason it is very important for researchers to pick up the understanding of the properties if these scales and their various influences on the responses gathered from their use.
More recent thought process highlights that after the questionnaire has been given to the respondents, all of the different features of that questionnaire become relevant information and have been seen getting used in judgement, interpretation and response to the different alternatives that are provided. Hence, like the constitution of words is an important aspect that depicts the quality of the questionnaire, other formal features or the symbols of non-verbal language are equally important aspects of question meaning to the respondents.
Generally the rating scales that are being seen as used in research are more balanced. They are composed of an equal, balanced number of favourable label that form an equal interval continuum that has been anchored by opposite poles, with or without the help of mid points. Findings have shown that unbalanced scales do not produce the same distribution results as the balanced scales. The position of the labels in the hierarchy of categories and the psychological meaning of the label descriptor has a strong influence on the way the respondents answer the questions. A positively unbalanced scale is seen as producing a much higher mean than a scale which is negatively balanced. Education research over the years has produced contrary results for balanced and unbalanced scales. In a study taken up by researcher in the United States the negatively packed scales where most of the scale points correspond to negative labels produced the highest mean rating of the instructors and on the other hand, in the case of positively packed scales where most of the scale points are corresponding to positive labels, the lowest mean rating is produced. Still there is scope in this field to investigate and extend research to understand the impact of balanced and unbalanced rating on the psychological properties of the scale, for example, reliability measures comparison for single or multi item measurement scales.