go to the School of Psychology home page Go to the UNE home page
Chapter 7 - Analysing the Data Part IV - Analysis of Variance Chapter 1 - Behavioural Science and research Chapter 2 - Research Design Chapter 3 - Collecting the Data Chapter 4 - Analysing the Data Part I - Descriptive Statistics Chapter 5 - Analysing the Data Part II - Inferential Statistics Chapter 6 - Analysing the Data Part III - Common Statistical Tests Sources of data Samples and populations Sampling methods Quantitative and qualitative approaches Questionnaire design Coding of data Readings and links


Chapter 3: Collecting the Data


Questionnaire design

A widely used aspect of research in the behavioural sciences is the self-report questionnaire. Personality traits, mood states, disorders, illnesses, abilities, recollections, and all sorts of other mental states are often assessed by self-report ‘pencil-and-paper’ tasks. As such these questionnaires constitute important research tools. They are in fact measuring tools for quantifying the amount of some psychological substance. We want to be able to say that someone has more of this substance (or construct) than someone else (a ‘between’ comparison) or we want to be able to say that one particular person has more of this substance (construct) at one time than they do at some other time (a ‘repeated measures’ comparison).

Measuring tools therefore need to be valid (i.e., measure what they are intended to measure) and reliable (i.e., get the same answer each time you are measuring the same thing). If our measuring tool is inaccurate than we cannot make precise claims about the psychological construct we are interested in because we cannot be confident that we know exactly how much of the construct is actually there. Once a psychological measuring instrument has been tested and improved and tested again in terms of reliability and validity it is considered standardised. A standardised test can then be used in a research project confident about the reliability and validity of the test as a measuring tool.

The development and evaluation of a new survey instrument can be a long and involved process. See Ray Chapter 13 for detail here. But essentially, the construct to be measured needs to be clearly defined and the questions or items chosen to measure the construct need to be carefully chosen and trialed on a few people for clarity and ambiguous wording. Next, the trial questionnaire needs to be given to a moderately sized sample and all the responses examined for how well they satisfy various criteria such as internal consistency (Cronbach’s alpha) or response variation.




© Copyright 2000 University of New England, Armidale, NSW, 2351. All rights reserved

UNE homepage Maintained by Dr Ian Price
Email: iprice@turing.une.edu.au