•  
  •  
 

Abstract

If, in the context of a survey, one wishes to know how old a participant is, one question suffices: “How old are you?” The same applies if one is interested in participants’ alcohol use per day, or whether the person is married or not. The use of a single item to measure constructs is also common practice in disparate fields such as job satisfaction ,1 stress research ,2 or happiness studies .3 In fact, how many different ways are there to ask how happy a person is? In education however, things are different. There, multi-item questionnaires and tests are traditionally the instruments with which constructs are measured. An important reason is that the measurement reliability of a single item cannot be estimated. One needs responses to more than one item to assess the internal consistency of a questionnaire or test. And sufficient reliability of an instrument is a precondition for it to be valid. However, here the same problem exists. Recently, I asked a group of medical students to fill in the eightitem self-efficacy† subscale of the well-known Motivated Strategies for Learning Questionnaire (MSLQ) developed by Pintrich.4 Afterwards, one of the students asked me: “Why do I have to answer largely the same question eight times?" Indeed, how many different ways are there to inquire about a student's self-efficacy? Despite the measurement concerns, single-item questionnaires are on the rise. In particular, in true experiments, researchers do not worry much about reliability issues. For instance, studies into cognitive load, experienced while working on a learning task, estimate with just one item the amount of cognitive load felt. In decision-making tasks, confidence in one's own judgement is also usually measured by a single item.

Download

Share

COinS