OVERVIEW
This assignment concerns an important step in data analysis: recoding. Recoding input variables (questionnaire results) is a routine task in SPSS. Recoding input variables is often used to change the codes for categories of a particular input variable or when there are too many variable input options for one question within a questionnaire. The original research design of the questionnaire may have warranted the inclusion of too many options. Often, the researcher is curious or interested in stating all of the variable input options for a given question or interested in past research endeavors that used the same scale as input variables.
When certain statistical calculations are run, the format of the input variable is non-negotiable. Most of the time, variables need to be combined when the number of particular responsesfor one category is too small to analyze. For example, a questionnaire could have asked respondents to give a specific age, which could be anywhere from 1–100 (100 choices). The original responses (continuous data) could be recoded into a new variable (ordinal data) that reflect 3–5 categories (30 and younger, 31–50, 51–70, and 71 and older).
Below is an example of recoding:
Original Code
New Codes
0 = never
1 = never to infrequently
1 = less than once a year
1 = never to infrequently
2 = about twice a year
1 = never to infrequently
3 = several times a year
1 = never to infrequently
4 = about once a month
2 = relatively frequently
5 = several times a month
2 = relatively frequently
6 = once day every week
2 = relatively frequently
7 = weekly
2 = relatively frequently
8 = several times a week
2 = relatively frequently
9 = no answer
9 = no answer
INSTRUCTIONS
To answer these questions, open up the Dell SPSS data set. If you have any questions about recoding, check out the websites below and the SPSS tutorials.
1. Recode the respondents based on total hours per week spent online into 2 groups: “5 hours or less (light users)” and “6–10 hours (medium users).” Calculate a frequency distribution.
2. Recode the respondents based on total hours per week spent online into 3 groups: “5 hours or less (light users),” “6–10 hours (medium users),” and “11 hours or more (heavy users).” Calculate a frequency distribution.
3. Form a new variable that denotes the total number of things that people have ever done online based on q2_1 to q2_7. Run a frequency distribution of the new variable and interpret the results. Note that the missing values for q2_1 to q2_7 are coded as 0.
4. Recode q4 (overall satisfaction) into 2 groups: “very satisfied” (rating of 1) and “somewhat satisfied or dissatisfied” (ratings of 2–4). Calculate a frequency distribution of the new variable and interpret the results. To interpret the results, run a frequency distribution of the original variable and compare and contrast it with the new frequency distribution of the new variable. Analyze the difference between these 2 frequency distributions along the dimensions of mean, standard deviation, range, kurtosis, and skewness. What do these differences mean?
5. Recode q5 (would recommend) into 2 groups: “definitely would recommend” (rating of 1) and “probably would or less likely to recommend” (ratings of 2–5). Calculate a frequency distribution of the new variable and interpret the results. To interpret the results, run a frequency distribution of the original variable and compare and contrast it with the new frequency distribution of the new variable. Analyze the difference between these 2 frequency distributions along the dimensions of mean, standard deviation, range, kurtosis, and skewness. What do these differences mean?
6. Recode q6 (likelihood of choosing Dell) into 2 groups: “definitely would choose” (rating of 1) and “probably would or less likely to choose” (ratings of 2–5). Calculate a frequency distribution of the new variable and interpret the results. To interpret the results, run a frequency distribution of the original variable and compare and contrast it with the new frequency distribution of the new variable. Analyze the difference between these 2 frequency distributions along the dimensions of mean, standard deviation, range, kurtosis, and skewness. What do these differences mean?
7. Recode q9_5 per into 3 groups: “definitely or probably would have purchased” (ratings of 1–2), “might or might not have purchased” (rating of 3), and “probably or definitely would not have purchased” (ratings of 4–5). Calculate a frequency distribution of the new variable and interpret the results. To interpret the results, run a frequency distribution of the original variable and compare and contrast it with the new frequency distribution of the new variable. Analyze the difference between these 2 frequency distributions along the dimensions of mean, standard deviation, range, kurtosis, and skewness. What do these differences mean?
8. Recode q9_10 per into 3 groups: “definitely or probably have purchased and might or might not have purchased” (ratings 1– 3), “probably would not have purchased” (rating of 4), and “definitely would not have purchased” (rating of 5). Calculate a frequency distribution of the new variable and interpret the results. To interpret the results, run a frequency distribution of the original variable and compare and contrast it with the new frequency distribution of the new variable. Analyze the difference between these 2 frequency distributions along the dimensions of mean, standard deviation, range, kurtosis, and skewness. What do these differences mean?
9. Recode the demographics as follows:
a. Combine the 2 lowest education (q11) categories into a single category. Thus, “some high school or less” and “high school graduate” will be combined into a single category labeled “high school graduate or less.”
b. Recode age (q12) into 4 new categories: “18–29,” “30–39,” “40–49,” and “50 or older.”
c. Combine the 2 lowest income (q13) categories into a single category labeled “Under $30,000.”
d. Calculate frequency distributions of the new variables and interpret the results. To interpret the results, run a frequency distribution of the original variable and compare and contrast it with the new frequency distribution of the new variable. Analyze the difference between these 2 frequency distributions along the dimensions of mean, standard deviation, range, kurtosis, and skewness. What do these differences mean?
Before continuing with the steps below, watch Presentation: SPSS Recode into New Variables, found in the Learn section for Module 6: Week 6.
To recode, go to: Transform => Recode into Different Variables => Select variable to recode => Select a Name and Label for the new variable => click Change => Old and New Values => Range Value Through Highest (= 5 for example) => New Value = 1 +> Add => then Range Value Through Lowest (= 6 for example) => New Value = 2 +> Add.
For frequency distribution: Analyze => Descriptive Statistics => Frequencies => Select Variable => Histogram and check Show Normal Curve on Histogram.
Including a Narrative in Your SPS Output File
You will note that some of the questions require you to interpret the results. SPSS allows you to include a narrative at any point in the Output file. To include a narrative, follow these simple steps:
Go to the point in the Output file that you wish to insert new text (highlight the point on the menu to the left), and then go to Insert => New Text.
Including any text in the Output file will reduce the amount of work you need to do in creating a Word document for the narrative. Consequently, you will only have to submit one file (the SPSS Output file) rather than multiple files.
Important: Do not restrict yourself to the steps listed. Have fun and experiment with options/layout/different stats options/graphs, etc. Read up on how to interpret results.
