How To Write a Good Survey for Psychological Research

//How To Write a Good Survey for Psychological Research
How To Write a Good Survey for Psychological Research 2021-04-18T21:46:59+00:00

psychology survey examplesSurveys are a common and powerful tool in psychological research. They’re an essential data collection tool that gathers self-report data in an experiment. Surveys follow two patterns: the questionnaire method and the structured interview. The questionnaire method is the more common of the two, with participants completing the survey without the researcher present. In a structured interview, the researcher is present and asks participants questions.

Using Surveys for Psychological Research

Researchers use surveys to investigate the opinions, behaviors, demographics and other characteristics of a group of people. The demographic information that’s collected can be brief or exhaustive, with information like ethnicity, sex, religion, political affiliation among the most commonly collected items. Hypothetical situations are also common survey questions, with researchers asking respondents what they would do in particular situations.

Surveys have a lot of advantages. They can be administered rapidly, with information collected over the phone, through the web, in person, or by mail.  They allow a great deal of data to be gathered, fairly rapidly. However, for surveys to work well, they must be constructed correctly. Surveys must have high reliability and validity. Reliability refers to a study’s ability to produce the same results across multiple administrations. In general terms, validity refers to a study’s ability to measure what it’s supposed to measure. Surveys can be invalid but produce reliable–although incorrect results. A valid survey will also produce reliable results. Reliability and validity are determined through intensive testing and statistical analysis. These procedures help make sure that surveys produce data that can be properly used in evaluating hypotheses.

Writing Great Psychology Surveys

Surveys, also called questionnaires, have many advantages, but they carry some potential problems too. People maybe be unduly influenced by the way questions are worded, the order in which response choices are presented, even the nature of the question itself. These extraneous influences limit the reliability and validity of a survey, which is why the best psychology surveys are crafted to avoid those issues.

To avoid these issues, make sure you eliminate biased language from your surveys. For example, the phrase “car wreck” may imply a serious collision, where “automobile accident” is much more neutral. Be aware of cultural and regional factors if you plan to make wide generalizations from your sampling.

All psychology experimentation is reliant on researchers having very clear ideas of what they’re investigating and how those questions are treated. The process of turning big-picture concepts into measures that can be tested, into questions that can be asked is called operationalization. Operationalizing a research question and all the survey items is an essential part of the process.

When it comes to wording, some essential rules to follow are:

  • Write each question in simple to read, easy to understand language. A good rule of thumb is to write questions at the 6th to 8th grade reading level.
  • Ask your question immediately, using unambiguous words. Be careful of words like “very,” “many,” “a lot,” and so forth.” They can introduce ambiguity.
  • Avoid dual, or double-barreled questions. A survey question that states “This product worked well and was easy to use” is an example of a double-barreled question, as there may be two very different possibilities for this single question.
  • Allow for a “does not apply” or “don’t know” response, but be aware sometimes they aren’t needed. For example,  “How easy was it making your purchase?” wouldn’t require a “does not apply” response option.

The importance of the questionnaire design and contents is utterly crucial. Each psychology survey question must focus on the variable or variables you’ve chosen to study. Survey design needs to be mindful of the following factors:

  • Surveys need to be fairly brief, taking no more than 15 minutes. The longer surveys are, the fewer people think carefully about their responses. They go faster in responding to get the survey finished as rapidly as possible.
  • Use a Likert-type scale as often as possible for measures of agreement, satisfaction and approval. Likert scales ranging from 5 to 7 points, with a middle “neutral” position. Likert scales allow for fine degrees of statistical analysis to be performed on data. Although there is occasional controversy about allowing middle positions (“slightly agree” or “slightly disagree”), research indicates that middle positions prevent over-polarized choices.
  • Keep coding consistent.  All survey responses are coded numerically for analysis.  All questions must be coded the same. That is, code responses so that the most positive outcome on each question is always given 5 points, for example, or however many points you choose. The critical point is that all survey questions have the same point value across questions. If “strongly disagree” on one question accrues a score of 1, all “strongly disagree” responses should be coded as a 1. This allows statistics to be generated on a survey rapidly, which is a big help in drawing conclusions.
  • Close-ended questions tend to be easier and faster to score and quantify than open-ended questions. Open-ended survey questions, on the other hand, give more complete information, yet require more complex analysis.

Steps in Carrying Out a Psychological Survey

  1. Define your variables. This is part of the aforementioned operationalization of your study. You have to have a clear and quantifiable definition for each variable in your survey.
  2. Develop a general hypothesis: Your hypothesis is a general idea about a testable situation. However, a hypothesis must be paired with a null hypothesis.  A null hypothesis is a prediction that the survey will yield no significant associations between your tested variables. If your study does show significant relationships, then you have rejected the null hypothesis. This becomes crucial in the statistical analysis phase, after collecting all your survey data.
  3. Perform a literature review. A literature review is a thorough investigation of all research about one’s hypothesis that has been published in scholarly outlets, like professional journals. A thorough literature review helps researchers avoid pitfalls while pointing out ways to improve one’s psychology survey topics.
  4. Design survey. A survey design must be tailored to the needs of the experiment, particularly the nature of the variables.
  5. Choose your participants. Who do you want to study and what do you want to investigate about your chosen population? A portion of a population is called a sample and choosing a sample is a topic of its own. The larger a sample is, the more accurately it reflects its population so that results can be generalized. However, it’s hard, if not impossible, to survey tens of thousands of people at one time. Sampling is a way to draw information that can then be analyzed by sophisticated statistical means to make statements about populations.
  6. Conduct the survey. Consider how you’re going to administer your survey. Phone surveys and internet-based surveys work differently from in-person interviews. In-person psychology surveys require a precise, standardized way of interacting with people taking the survey.
  7. Analyze results. The first pass through a data set is conducted by statistics programs. Interesting relationships may then be highlighted and examined. It’s important to note that a single survey doesn’t have a tremendous amount of generalizability. If a study is investigation a broad demographic, for example political standings among African-American men from age 18 to 25, and one’s sample size is 500, it’s going to be problematic to attempt to apply findings from a small sample size.

Surveys work best when they are thoughtfully constructed and given to large sample sizes. By paying attention when the psychology survey topics are chosen and during the survey’s design phase, researchers can build an instrument that will yield data that’s reliable, valid and generalizable.

Clifton Stamp

B.S. Psychology | Arkansas State University

M.A. Rehabilitation Counseling | Arkansas State University

M.A. English | Arkansas State University

November 2019

More Articles of Interest: