Statistics for Research - L4 - Introduction to SPSS and Data Entry

SPSS data analysis starts with a disciplined setup: define every variable first, then enter responses using numeric coding, and finally verify that the dataset matches the questionnaire. In social science research, anything measured—age, gender, education, university, job details, and survey items—counts as a variable, and SPSS requires those variables to be specified in “Variable View” before any data go into “Data View.” The workflow matters because SPSS treats analysis as number-based, so categories and Likert-scale answers must be converted into consistent numeric representations.

A typical study setup begins with demographic variables (e.g., age, gender, education, program, university, job rank, tenure) and then moves to constructs measured indirectly through indicators. For instance, an unobserved “latent” construct (like internal stakeholder responsibilities) is operationalized through multiple observed items (six indicators in the example). Each indicator is measured on a five-point Likert scale where 1 corresponds to “strongly disagree” and 5 to “strongly agree.” Since respondents select only one option per item, the dataset should contain single numeric values per indicator per respondent—no multi-select entries.

In “Variable View,” the type and measurement level for each variable must be set. Age is treated as interval/ratio (numeric with meaningful ordering), gender as nominal (no inherent order), education as ordinal (ordered categories such as degrees), and university as nominal. For non-numeric categories, SPSS uses value labels. The example shows mapping gender categories to numbers (e.g., 1 = male, 2 = female) and similarly encoding education levels (e.g., 1 = Masters, 2 = MPhil/MS, 3 = PhD, etc.) and universities as numeric codes. Likert items can reuse the same numeric scale (1–5) with labels for each response option.

Once variables are defined, the next step is entering data in “Data View.” Each row represents a respondent, and each column corresponds to a variable or indicator. When questionnaires exist in print, the dataset should include a questionnaire number so each row can be traced back to the correct paper. The example demonstrates entering respondent #1 with values for demographics and six responsibility indicators, then repeating for respondent #2 and onward.

After data entry, SPSS needs a quick correctness check. A practical first validation uses “Analyze” → “Descriptive Statistics” → “Frequencies,” including minimum and maximum checks. If a value falls outside the expected range—such as a Likert item showing a 6 when the scale runs from 1 to 5—SPSS highlights the offending variable. The fix is to locate the exact cell (using “Edit” → “Find” to search for the invalid value), then cross-check the corresponding questionnaire row to correct the entry. This cycle—define, code, enter, and validate—sets up the dataset for later analysis steps.