CERG

The Evaluation Resource Centre

Home
Home | Site Map | Design & Development | Survival Kit | Ethics | Bibliography | Glossary | Doing An Evaluation
Home
     

> Design & Development > Analysing > Content Analysis

Analysis Methods

 

Content Analysis

Content analysis is one of the basic tools of the qualitative researcher. It is the simplest method for determining what is in the qualitative data you have collected.

For the sake of simplicity we will be talking about a single qualitative question. For example, you may have asked your respondents to write down how they felt at the end of the exam they had just completed. But most research will have a number of questions and each question has to be dealt with using the process outlined below.

Step 1

Read through a sample of the respondents. If you your total sample is small, about 20 or less, then you would read them all. Otherwise take about 20% of them.

Get a feel for the what the respondents are saying. Is it about positive and negative feelings? Is it about what they tried to write?

Write down the content themes that are appearing. When you think you have a new theme but you think it could be similar to a previous them, write it down. It is better to start with too many themes.

Step 2

Review what you have found from the first step.

Look at the content themes and decide whether any of them can now be combined. In thinking over what you have found, do not reduce the 'richness' of the data in favour of convenience.

If you are not clear on what you have, then go back and re-read some of the questions.

Step 3

Now you are ready to organise all responses against the content themes.

You can group them according to the major themes and explore what this all adds up to against your reason for collecting the data.

As you are working with the rest of the responses you may find that you have to add to your content themes. There is nothing wrong with this but usually a 20-25% will cover the effective content.

Step 3 alt

Some researchers would code each of the responses against the content themes that had been found. For example, each theme might be given a letter value and each respondent would be coded to show whether or not he/she used that theme.

In using a coding system, the researcher has the possibility of using nonparametric statistics to analyse their data.

Comment

There are a number of software tools around for carrying out various forms of thematic analysis of qualitative data. Some of these are excellent if:

  • you have the time to learn to use them
  • you have the time or money to have your data transcribed
  • you have enough data to warrant all of this

 

Home
Copyright © 2004 Monash CERG