Sequence analysis process
• Make a record of the raw data
• Data ready for counting events and transitions
• Count events using ‘Find and Replace’ in Word. (See picture below)
• Pick out the most frequent 15-ish, and put the others into groups, eg ‘Low frequency events’. Recode the data accordingly into a new file.
• Count all transitions between these new event categories using ‘Find and Replace’ in Word.
• Make the Transitional Frequency Table, including 'Standard Normal Residuals'. (See pictures at the end for details of the steps, using SPSS.)
• Logic of using Standard Normal Residuals
• Chi-squared table
• Table of Standard Normal Residuals
• Pick a criterion for SNR (2 is often about right), and identify the cells that meet that criterion – the ‘hot spots’ in the table. These will then be the transitions that appear in your results / transition map.
• Link the transitions (event pairs) together
• Software for drawing transition diagrams
http://www.yworks.com/en/products_yed_about.html
• The end product (a drink-driving study)
** Making the transition table in SPSS: a step-by-step guide **