The Code Book
A code book is a compilation of the codes, and is to be used as a reference tool. In addition to setting up the codes themselves, code books can provide information on the definition of a particular variable to aid in assigning the proper code. Because coding is subjective, there can be some variations in how codes are applied by different coders. Code books aren't just used for surveys with open-ended questions; they're used for closed-end, quantitative responses as well.
Here are the basic elements of a code book:
- Variable names
- What the variable stands for (i.e., its label)
- How the variable was measured (i.e., ordinal, nominal, scale)
- How the variable was recorded in the data set (field width, number of decimal places)
- If it was a scale, the units of measurement
- For categorical variables, if they were coded numerically, the list of codes and what they represent
Code books can be set up manually in Word or Excel. When you set up a code book in Excel, there may be separate worksheets for variable labels (i.e., ethnic origin) and values labels (i.e., Caucasian). SPSS is another popular program for this type of analysis. In SPSS, you can automatically generate a code book from the data set. Here's what a variable information table would look like from an SPSS codebook:
Now, here's a table showing the variable values:
 |
Scoring the Data
Once the codes have been applied to the data set, the data can be put through analysis. There are software packages you can use to apply codes to the text in order to build models and help in data management known as CAQDAS (Computer Assisted Qualitative Data Analysis Software). For example, they can indicate the frequency of particular answers to a question. Many of these tools are free, and you can find them by searching for ''free qualitative data analysis software''.
When phrases, words, and themes have been turned into quantitative data and put into spreadsheets, they can be evaluated via cross-tabulation to look for relationships between variables. You can also look for frequency distribution of variables - how often they occur within the sample. Most spreadsheets and statistical analysis packages also create charts and graphs that can be used for presentation of the data.
Lesson Summary
Content analysis is the process of sifting through answers to open-ended questions and looking for patterns and themes. Through coding, words and phrases that are used frequently can be assigned alphanumeric codes. Once codes have been defined, code books are created that compile the codes so that they can be used as a reference tool. The code book can be electronic form in Word or Excel, or there are also packages like SPSS that can automatically generate a code book from the data set.
There are also software packages that can analyze the survey data and help build models. These are known as CAQDAS (Computer Assisted Qualitative Data Analysis Software). Once codes have been applied to data and the data is in tabular form, the results can be analyzed through techniques like cross-tabulation - looking for relations between variables - or frequency distribution - looking at how often values occur within the sample.
