Blog Post#6: Writing Effective Survey Questions
Hi everyone! I just got off reading through the handout 'Writing Good Survey Questions: Tips and Advice' and the article '8 Types of Survey Questions to Get You All the Data You Need' today and I could not wait to share my first thoughts about it! The author, Nemanja Jovancic in his article mentions that deciding on the exact goals of our survey beforehand makes the choice of an adequate survey question type much easier. Some of the factors that may determine the type of survey questions we are going to use include the type of information we need, the depth of information we need, and the amount of time our respondents have available.
The author further highlights that sometimes just slightly varying the types of survey questions and answer choices can have a significant impact on the value and quality of the results we obtain, as well as on the response rate. Thus, it is essential to pay special attention to the length of our survey and the number (not just the type) of questions included. Questions in our survey should be short, i.e., one or two lines of text the most. It is advisable for our survey to have 8-10 questions.
The author suggests including 8 types of survey questions in our survey for covering all our data requirement. These include Dichotomous (Yes/No) questions, multiple choice questions, checkbox questions, rating scale questions, likert scale questions, matrix questions, open ended questions and demographic questions.
Some of the general tips for writing good survey questions include using clear wording (no abbreviations), avoiding leading questions and making questions specific. While wording our responses, it is advisable to use balanced scales (i.e., equal number of positive and negative response options), make the response alternatives exhaustive and mutually exclusive, etc.
In fact, I came across something similar while reading about a chapter on 'The Experimental Evidence' from the book 'Survey Questions: Handcrafting the Standardized Questionnaire' by Jean M Converse and Stanley Presser. In the chapter, one of the important things the authors explain is that particularly with general questions, the same question can sometimes mean different things to different people. Thus, specific wording is more appropriate to confirm uniform meaning among the people who are being surveyed. Further, specificity aids respondent recall. The authors also cite an experiment in the chapter to explain the advantages of specificity. However, general questions can't be ruled out entirely. They have their uses when there is not the time to ask about everything in specific detail and also to get to know about people's opinion. Thus, it is essential to include most of the types of survey questions for covering all the data we require. Here is the link to this book:
There is also an interesting video on this topic that provides some tips on developing survey questions: