Home Computer Science OECD guidelines on measuring subjective well-being.
The number of response options to offer
Sensitivity or discriminatory power is an important factor in scale design. A good subjective well-being measure will be one that enables variability between respondents to be detected where it exists. In addition to the way in which a question is worded, the number of response options offered is a critical determinant of scale sensitivity - and this is particularly so for measures that rely on just one question, rather than summing across a number of items.
Offering too few response options may not enable some respondents to fully express themselves. This can cause meaningful variations in scores to be lost, and respondents may also grow frustrated, leading to lower quality or even random responding. On the other hand, too many response categories can potentially increase cognitive burdens, especially if the response options presented offer finer distinctions than respondents are able to make in their own judgments or on the basis of recall from memory. Thus, offering too many response options also has the potential to demotivate respondents, leading to lower-quality data.
There is, therefore, a trade-off between capturing as much meaningful variation as possible on the one hand, and minimising respondent burden and frustration on the other. Survey mode also has important consequences for the cognitive burden of different response formats, something that will be discussed in more detail later. The preferred number of response options may also be different for different types of subjective well-being measures, given the differences in the underlying constructs. Finally, respondents may vary in their ability to cope with cognitive burdens and in their preferences for simple versus complex response options. Thus a compromise must be struck between the needs of the data users, the respondents being surveyed, and the constraints of the survey method.
|< Prev||CONTENTS||Next >|