Google Forms Survey: Creating a Form and Collecting Thesis Data

What a Google Forms survey means in quantitative research

In quantitative research, a Google Forms survey usually means using Google Forms to create an online questionnaire, sending the link to respondents, and exporting the data to Google Sheets or Excel. You can then import this file into SPSS as .xlsx, .csv, or convert it into .sav format.

You should separate the tool from the method. Google Forms only displays the questionnaire and records responses. The quality of your study still depends on the research model, indicators, respondent selection, sample size, and data-cleaning process. A polished form with items that do not measure the intended construct still produces data that is difficult to defend.

For a thesis, a sensible process is usually to define the objectives and hypotheses, select the scales, write the questionnaire, pilot it with a small group, distribute the form to the right respondents, check the responses, and only then import the data into SPSS. If you are still unclear about the difference between a questionnaire and a survey, read what is a questionnaire before opening Google Forms.

Google Forms works well when you need a tool that is quick to deploy, free, and easy to share. You can also use DoThesis's survey form to design a more structured data-collection process. Whichever tool you use, save the questionnaire versions, collection dates, and response-exclusion criteria so you can explain your decisions to your supervisor.

How to determine the sample size for a Google Forms survey

The sample size is the number of valid responses remaining after you remove incomplete answers, implausibly fast responses, or responses that appear to have been completed without attention. Do not treat the number of people who clicked the form as your official sample size. The number you report in your thesis is the number of valid questionnaires included in the analysis.

If the population is finite and you know its size, you can refer to Yamane's formula:

n = N / (1 + N e²)

Here, N is the population size, e is the allowable error, and n is the required sample size. This approach is commonly cited as (Yamane, 1967). When you are studying a proportion in a population, you can refer to the sample-size formula for proportions in (Cochran, 1977). Do not use Yamane's formula before deciding which population you are assuming.

If you run a regression with m independent variables, a commonly used rule is n at least 50 + 8m, following (Tabachnick and Fidell, 2013). For factor analysis, Hair et al. discuss the rule of 5 to 10 observations per item in (Hair et al., 2010). These are reference points for planning. They are not a reason to choose an arbitrary number.

The figures in the example column illustrate the calculation. They are not a mandatory sample size for your topic. If you expect many responses to be excluded, distribute the form to more people than the minimum. For example, if you need 250 valid questionnaires and expect 15% to be unusable, you should aim to collect approximately 295 responses.

In Chapter 3, state the population, sampling method, collection period, number of responses received, and number of valid responses. If you use convenience sampling through social media, do not describe it as random sampling. These two methods have different conditions and different levels of representativeness.

Designing a questionnaire for a Google Forms survey

Before creating the form, prepare a variable codebook. Each latent variable should be linked to its indicators and an appropriate scale source. A quantitative questionnaire usually includes an introduction, screening questions, descriptive questions, and groups of measurement statements.

Writing the introduction and screening questions

The introduction should state the academic purpose, estimated completion time, eligible respondents, and confidentiality principles. Do not promise that responses are completely anonymous if the form collects email addresses or requires respondents to sign in.

Place screening questions at the beginning of the form. For example, if you are studying purchasing behavior on a platform, ask whether the respondent has used that platform during the study period. If the answer is “no,” Google Forms can send the respondent to the ending section through the option that directs respondents to a section based on their answer.

Choosing question types in Google Forms

Use Multiple choice when a question has only one answer, Checkboxes when respondents can select several answers, Dropdown when the list is long, and Linear scale for a simple rating. For scales with several items, Multiple choice grid or a separate question for each statement usually gives you better control over the data.

Avoid placing too many items in one matrix when most respondents use a phone. Check the required-response settings as well. Making every question mandatory can cause respondents to select random answers when a question does not apply to them.

Using a consistent Likert scale

A Likert scale measures agreement with a statement rather than asking generally how respondents “feel.” The technique originates from (Likert, 1932). You can use 5 or 7 points, but label the scale clearly from the lowest to the highest level and use it consistently across the entire scale group.

For example, the statement “I can easily find the information I need on the application” could use values from 1, completely disagree, to 5, completely agree. Do not mix agreement questions with frequency questions in the same matrix unless the instructions make the difference clear.

If an item is reverse-coded, mark it in the codebook and process it before running Cronbach's Alpha. Do not write an unnecessarily confusing reverse-worded statement just to create variety. Respondents using a phone can easily miss a negative word, and you may later have difficulty explaining the data.

Distributing the Google Forms survey and collecting data

After completing the form, send it to a small group to check spelling, links, section logic, completion time, and the way the data are exported. This checks the tool. It does not replace evaluating the scale with the official dataset.

Google Forms is easy to share through a link, turn into a QR code, and synchronize with Google Sheets. However, you should check editing permissions, version history, and the setting that limits each person to one response before publishing the form. If respondents can edit their answers after submission, record the point at which the dataset is finalized.

DoThesis has fillform.info, our form tool for collecting Vietnamese questionnaire responses. You can use it when you need to manage the form according to your own data-collection process. Google Forms remains a suitable alternative when you want a familiar tool and do not want to change platforms.

When distributing the form, record each distribution source. For example, you may share it in a student group, a professional community, a class email list, or social media. If respondents from each source have different characteristics, store that information in a separate column or tracking file. Do not force respondents to select a respondent group if you have not actually asked them about it.

Set a clear collection period, such as a start date and a closing date. More responses do not automatically mean better data. If one group accounts for too much of the sample, your data may be unbalanced relative to the target population. Track the sample structure using descriptive variables such as gender, age, occupation, or usage experience when those variables fit your topic.

Cleaning data before importing it into SPSS

In Google Forms, open the Responses section and export the data to Google Sheets or download the Excel file. Before importing it into SPSS, create a read-only original copy. Work on a duplicate and name each version, such as khaosat_v1_raw and khaosat_v2_clean, so you can return to an earlier version when you find an error.

Review columns such as Timestamp, email, name, or response code according to the purpose of your study. Timestamp helps you check completion time, but it is usually not an item entered into Cronbach's Alpha. Email addresses and names should usually be excluded from the analysis file unless they are necessary. Keep consent confirmations, screening questions, and demographic variables when they support the sample criteria or sample description.

Check for the following signs:

• A response leaves many items blank or fails a screening question.
• The completion time is too short compared with a reasonable pilot completion time.
• One person selects the same value for every statement in a group, a pattern commonly called straight-lining.
• The answers are inconsistent, such as selecting “never used it” and then rating several features in detail.
• Values fall outside the permitted range, such as a value of 0 or 6 in a scale from 1 to 5.
• Duplicate rows appear when the form allows multiple submissions but the study requires one response per person.

Do not delete data simply because one response lowers Cronbach's Alpha. First check whether the exclusion criteria were written in your research method. Keep a tracking table containing the response code, exclusion reason, and processing date. When importing the data into SPSS, use short variable names with no accents and no spaces, such as PU1, PU2, and SAT1. You can write the full Vietnamese wording in the variable labels.

If you use a Likert scale from 1 to 5, check that every value falls within this range. For a reverse-coded item, the usual conversion formula for a 1 to 5 scale is 6 - old value. Perform this conversion only after confirming the item's direction in the codebook. Then save the .sav file and keep the raw data file unchanged for comparison.

How to present a Google Forms survey in Chapter 3

In Chapter 3, present the process from the study population to the data-collection instrument. State the scale sources, how you adapted the wording to the research context, the questionnaire structure, the scales used, how you distributed the form, the collection period, and the data-cleaning criteria.

The sample-size section must match the results section. If Chapter 3 says that you expected 300 questionnaires but Chapter 4 analyzes 247 valid questionnaires, explain how many questionnaires were distributed, how many were returned, and how many were excluded. Do not change a number between chapters simply to make a table look cleaner.

How to write this in your thesis: “Data were collected through an online questionnaire using [Google Forms or the tool used], from [start date] to [end date]. After [number] responses that did not meet the [exclusion criterion] were removed, [number] valid responses remained for analysis in SPSS. The questionnaire consisted of [number] sections, with the item groups measured using a [5 or 7]-point Likert scale.”

If you use a sample-size formula, include the formula and source in the explanation. For example, the rule of 5 to 10 observations per item is cited as (Hair et al., 2010). If you determine the sample size from the number of independent variables in a regression, state the number of variables m and show the calculation according to (Tabachnick and Fidell, 2013).

The questionnaire, coding, and data file should be consistent. You can also read survey questionnaire or Likert scale to review the structure before closing the form.

Common errors when using a Google Forms survey

The first error is opening the form before finalizing the model. Once hundreds of responses have been collected, adding a new variable or changing the wording means that data from different stages are no longer consistent.

The second error is using leading questions, such as describing a product as “very convenient” and then asking respondents whether they find it convenient. The wording should be neutral and measure one idea at a time. The statement “The application is easy to use and processes requests quickly” contains two ideas, and a respondent may rate the two parts differently.

The third error is sharing the form with the wrong respondents. Sending a questionnaire about business customers to a group of people who have never used the service will produce many invalid responses. A screening question only helps you check eligibility. It cannot repair a poorly planned distribution strategy.

The fourth error is allowing email addresses, names, or phone numbers to remain in the analysis file when they are unnecessary. Collect identifying information only when there is a clear reason, restrict access, and separate the contact file from the SPSS file.

Finally, many students download the file from Google Sheets straight into SPSS without checking the data types. Review variable names, value labels, missing-value codes, reverse-coded items, and out-of-range values before running Cronbach's Alpha or EFA. If the form has faulty logic, manually correcting the data without recording the history will make it difficult to explain your decisions when asked.

Frequently asked questions

What is a Google Forms survey, and can I use one for a thesis?

A Google Forms survey uses Google Forms to distribute an online questionnaire and collect responses. You can use it for a quantitative thesis if you define the target respondents, sample size, scales, and data-cleaning criteria correctly.

Does a Google survey have a respondent limit?

Check the policy and account type available when you launch the survey because limits can change. For a thesis, the more common problems are invalid data, an unbalanced sample, or distribution to the wrong respondents rather than the number of responses alone.

Should I use Google Forms or another survey tool?

Google Forms is suitable when you need to create a form quickly, share it easily, and export the data. DoThesis's fillform.info is suitable when you want a Vietnamese form tool that follows your own data-collection process. Choose based on data-management permissions, response controls, and how you will export the file to SPSS.

How many responses does a Google Forms survey need?

The sample size depends on the population, number of items, number of independent variables, and analysis method. You can refer to the formula in (Yamane, 1967), the rule of 5 to 10 observations per item in (Hair et al., 2010), or the regression rule 50 + 8m in (Tabachnick and Fidell, 2013). Report the number of valid questionnaires, not only the number of people who opened the form.

What should I do if the survey form has missing data?

First check which questions are optional, which questions have faulty section logic, and whether the respondent belongs to the target population. During cleaning, save the original dataset and record the response code and exclusion reason. Do not fill missing values with a convenient number unless you have a clear methodological rule for doing so.

Open your Google Forms file and check the codebook, screening questions, and valid sample size before exporting .csv or .xlsx. If you need to run the analysis on your own data file, use M4 data analysis to continue with SPSS and SmartPLS.