You collected 400 responses in two weeks, opened the spreadsheet, and found words instead of numbers, a checkbox question crammed into one column, and three people who finished a 40-item survey in under a minute. These are the standard problems of online forms, and all of them are solved in about an hour if the steps are taken in the right order. This Google Forms to SPSS guide takes you from the Responses tab to an analysis-ready SPSS file: exporting, tidying the sheet, importing with the right variable types and labels, recoding Likert text into numbers in the correct order, reverse-scoring, computing scale scores, and screening duplicates and careless responses, with a worked example at the end.
This article is part of our guide How to Write a Master’s Thesis by the Mastermind PhD team and relates to our Statistical Analysis service. Sources last checked: November 2026.
What is the Google Forms to SPSS workflow? It is the sequence that turns responses collected by a Google Form into an SPSS data file ready for analysis: exporting the response sheet, shortening headers and splitting multi-select questions, importing into SPSS with correct variable types, assigning labels and missing-value codes, recoding text answers into numeric codes in the intended order, reversing negatively worded items, computing scale scores, and screening for duplicates and careless responding. The summary charts that Google Forms displays are not an analysis a thesis can use.
Step 1: Close the form and export the responses
Start every Google Forms to SPSS export the same way: switch off “Accepting responses” first, so the number of responses you export is the number you report. Google’s help page for form responses describes two export routes: open the Responses tab and click “View in Sheets” to send everything to a linked Google Sheet, or click More and “Download responses (.csv)”. The CSV downloaded from the form contains the raw submissions only; any cleaning you do in the sheet is not in it, so if you tidy data in Sheets, export from Sheets with File, then Download, then Microsoft Excel (.xlsx) or CSV.
Prefer .xlsx when answers contain non-Latin characters (Arabic, accented letters), because a CSV opened carelessly in Excel loses the encoding. Before any Google Forms to SPSS conversion, keep the exported original untouched in a separate folder, work on a copy, and note the export date and the response count for the methodology chapter.
Step 2: Clean the spreadsheet before the Google Forms to SPSS import
- Replace the long question headers with short variable names that SPSS accepts: letters, digits and underscores, beginning with a letter, no spaces (Q1, Q2, Gender, Exp). Keep the full question text in a codebook; SPSS also stores the original header as the variable label when it imports from Excel.
- Keep the Timestamp column for screening; delete email addresses and names after the duplicate check to protect respondents.
- Split checkbox questions. A “select all that apply” question exports as one cell such as “Internet, Friends, Lecturers”. Create one column per option with a formula like =IF(ISNUMBER(SEARCH(“Internet”, C2)), 1, 0), so each option becomes a 0/1 variable.
- Delete empty rows and the test submissions you made while building the form.
- Change values with formulas or find-and-replace you can document, never by retyping.
📱 Ask a statistician about your data — Send a short description of your instrument and sample size on WhatsApp and get a suitable analysis plan. Every analysis comes with the raw SPSS file and a results report.
Step 3: Import into SPSS and set the variable properties
For Excel files: File, then Import Data, then Excel; choose the worksheet, tick “Read variable names from first row of data”, leave the percentage of values that determines data type at its default, and click OK. For CSV files: File, then Import Data, then CSV Data; confirm that the first line contains variable names, that the delimiter is a comma and the text qualifier a double quote, and check the decimal symbol. Then open Variable View and complete what a Google Forms to SPSS import cannot know: Type (numeric or string), Decimals (0 for codes), Label, Values, Missing and Measure.
| Google Forms question type | How it exports | What to do in SPSS |
|---|---|---|
| Multiple choice used as a Likert item | Text labels (“Agree”) | Recode to numbers in scale order (Step 4); Measure: Ordinal or Scale for scale scores |
| Linear scale | Numbers 1 to 5 | Already numeric; add value labels and check the direction |
| Multiple-choice grid | One column per row, text labels | Recode all rows in one run |
| Checkboxes | One cell with comma-separated choices | Split into 0/1 variables in the sheet; define a Multiple Response Set |
| Dropdown (nominal categories) | Text | Automatic Recode is acceptable because order does not matter; or assign codes manually |
| Short answer (age, years of service) | Text or number | Convert to numeric; clean entries such as “25 years” |
| Paragraph | Free text | Keep as string for qualitative coding |
| Timestamp | Date and time | Keep as a date variable for screening |
Step 4: Recode Likert text into numbers in the right order
In a Google Forms to SPSS export, multiple-choice items store the answer as text. The most damaging shortcut is Transform, then Automatic Recode, because it numbers categories alphabetically: Agree becomes 1, Disagree 2, Neutral 3, Strongly agree 4 and Strongly disagree 5. The scale is now meaningless, and the first symptom is a Cronbach’s alpha near zero or negative.
The correct route is Transform, then Recode into Different Variables. Select every item that shares the same response anchors, give the new variables names with a suffix (Q1n, Q2n and so on), and in Old and New Values enter the pairs: “Strongly disagree” to 1, “Disagree” to 2, “Neutral” to 3, “Agree” to 4, “Strongly agree” to 5. Copy the labels from the data itself so that apostrophes, spaces and capitals match exactly; a single mismatch produces system-missing values.
Click Paste to keep the syntax, which for twelve items is one statement: RECODE Q1 TO Q12 (‘Strongly disagree’=1) (‘Disagree’=2) (‘Neutral’=3) (‘Agree’=4) (‘Strongly agree’=5) INTO Q1n TO Q12n. Add value labels to the new variables, set Measure, and check one item with a crosstabulation of old against new values. For future surveys, use the Linear scale question type, which exports numbers from the start.
Step 5: Reverse-score negatively worded items
Items written in the opposite direction (“The course was a waste of time” in a satisfaction scale) are reversed after recoding so that a high score always means more of the construct. For a 5-point scale compute 6 minus the item; for a 7-point scale, 8 minus the item. Use Transform, then Compute Variable, write the result to a new variable with the suffix _R and keep the original. Reliability Analysis is the check: an item that still shows a negative corrected item-total correlation has not been reversed.
Step 6: Compute scale scores
Use Transform, then Compute Variable, and the MEAN function with a minimum number of valid items: JobSat = MEAN.6(Q1n, Q2n, Q3n, Q4n, Q5n_R, Q6n, Q7n, Q8n) returns a score only when at least six of the eight items were answered. Averages keep the score on the original 1-to-5 metric, which makes interpretation keys usable; sums are appropriate when the instrument’s norms are expressed as sums. Label the new variable, run Descriptives to confirm the minimum and maximum fall inside the scale range, and state the missing-item rule in the methodology chapter.
Step 7: Screen duplicates, speeders and straightliners
- Duplicates. Data, then Identify Duplicate Cases. Match on email or participant code if you collected one; otherwise on the full set of item variables. Two identical response patterns submitted within minutes of each other are one respondent.
- Speeders. Google Forms records the submission time, not the start time, so completion duration is unavailable unless you added a timing add-on. Compare the timestamps of consecutive submissions from the same channel and treat implausibly fast runs as suspect.
- Straightliners. Compute the standard deviation of each respondent’s answers across the scale items (Compute Variable with SD(Q1n TO Q28n)). A value of zero across 28 items that include reversed statements is a respondent who clicked the same column throughout; exclude and document.
- Out-of-range and missing values. Run Frequencies on every item, and count missing values per case (NMISS) to apply the exclusion rule you stated, for example more than 20% of items missing.
Worked example: Google Forms to SPSS, 412 responses to a clean file in one hour
A student-satisfaction survey with four subscales and 28 items returned 412 Google Forms responses. The sheet was exported as .xlsx. Screening removed 9 duplicates (same email, identical answers, two minutes apart), 14 straightliners (SD = 0 across the 28 items, two of which were reversed statements) and 6 cases with more than 20% of items missing, leaving 383 valid responses (93 percent of submissions), figures that went straight into the methodology chapter. Headers were shortened to Q1 to Q28 plus Gender, Year and Faculty, and the checkbox question on information sources was split into six 0/1 variables before import.
The Likert text was recoded for all 28 items in one RECODE statement, items Q7 and Q19 were reversed, and the four scale scores were computed with MEAN.6 or MEAN.5 depending on subscale length. The reliability check showed the value of the order of steps: the teaching-quality subscale returned Cronbach’s α = .62 before Q7 was reversed and α = .84 afterwards.
Descriptives confirmed the scores stayed within range (teaching quality M = 3.62, SD = 0.71, skewness −0.38), and the first inferential test ran cleanly: satisfaction did not differ significantly by gender, t(381) = 1.94, p = .053, d = 0.20. Every step was saved as syntax and placed in an appendix, so the full Google Forms to SPSS preparation can be rerun from the raw export in seconds.
Which Google Forms to SPSS mistakes cause problems later?
- Analyzing from the Forms summary charts instead of the exported data.
- Automatic Recode applied to Likert items.
- Deleting or overwriting raw columns and keeping no codebook.
- Importing a CSV with the wrong encoding so non-Latin text appears as symbols.
- Leaving “Prefer not to say” coded as a scale point rather than as a user-missing value.
- Treating a checkbox column as one nominal variable.
- Not reporting how many responses were excluded and why.
- Testing hypotheses on single items instead of scale scores.
📱 Ask on WhatsApp now — 5 years of experience across all disciplines and 14 academic services. Two-stage payment (80/20); revisions are free until supervisor approval, within 30% of the chapter.
Frequently asked questions
Should I export as CSV or Excel?
Excel (.xlsx) when any answer contains non-Latin characters, because it preserves the encoding; CSV is fine for purely Latin text. Remember that the CSV downloaded directly from the form excludes any edits made in the linked sheet.
Can SPSS read a Google Sheet directly?
No. For Google Forms to SPSS, download the sheet as .xlsx or CSV and import the file. If responses are still arriving, close the form first so the file and the methodology chapter report the same count.
How do I handle “Prefer not to answer” options?
Give them a code outside the scale range (for example 99), declare it as a user-missing value in Variable View, and report how many respondents chose it. Do not leave it as a sixth scale point.
Should I make every question required to avoid missing data?
Requiring the scale items reduces missing values but can push reluctant respondents into random answers or abandonment. A common compromise is to require the scale items and leave sensitive demographic questions optional with a “prefer not to say” option coded as missing.
My form used the Linear scale type. Do I still need to recode?
No recoding is needed because the values are already numeric, but you still add value labels, set the measurement level, reverse negatively worded items and compute scale scores. Check that 1 and 5 mean what the instrument intended.
Sources
- Google. View & manage form responses (Google Forms Help). https://support.google.com/docs/answer/139706 (accessed October 2026)
- IBM. IBM SPSS Statistics documentation: Reading Excel files; Text Import Wizard; Recode into Different Variables; Identify Duplicate Cases; Multiple Response Sets. https://www.ibm.com/docs/en/spss-statistics
- Pallant, J. (2020). SPSS Survival Manual (7th ed.). Open University Press.
- Field, A. (2018). Discovering Statistics Using IBM SPSS Statistics (5th ed.). SAGE.
- Dillman, D. A., Smyth, J. D., & Christian, L. M. (2014). Internet, Phone, Mail, and Mixed-Mode Surveys: The Tailored Design Method (4th ed.). Wiley.
Sources last checked: November 2026. This article is general educational guidance and does not replace your university regulations or your supervisor’s instructions. How we prepare our content.