Presentation on theme: "Surveys and Questionnaires. How Many People Should I Ask? Ask a lot of people many short questions: Yes/No Likert Scale 5 4 3 2 1 Ask a smaller number."— Presentation transcript:
Surveys and Questionnaires
How Many People Should I Ask? Ask a lot of people many short questions: Yes/No Likert Scale Ask a smaller number of people some longer questions: Open ended – “What is your opinion of….?” “How would you solve this problem?”
How Many People Should I Ask? You can combine yes/no or Likert scale with a few open-ended questions. This will give good information with some details. OR: Interview one person. Summary: Many people, short answers. Few people, long answers.
It all depends It depends on your topic. Gather enough information to write a report about your results. Too little information = nothing to say Too much information = you have to write a book!
What do I ask? Basic data to describe your participants – Age and gender – Area of study – Anything related to your topic (Does their family live in San Juan or on a farm?) We had ten participants. They were all between the ages of 19 and 21, three women and seven men. Most were biology majors but two were in business administration.
Your Topic FIRST, put a title or heading saying what the questionnaire is about, who it is from, etc. After the general data (Demographic Data), ask your questions. You can put the age/gender/etc. at the end if you like. Some researchers prefer that but we are not asking for any sensitive data (anything about $$ is sensitive). Do not ask any questions that you do not need to know. Ask only what is 100% necessary You must analyze every question in detail so think about what you need to know and why you need to know it. LAST, write THANK YOU at the bottom.
The Questionnaire Do not crowd the page Do not ask too many questions (so the person feels tired just looking at it, like the UPRM teacher evaluation form) Do not use big fancy words or complicated grammar Have one question per item (do not combine questions – people will not answer them) Do not have “leading questions” Do not have bias
Bias Bias is your personal opinion. You do not want the participants to know your personal opinion. Example: Via Verde is the stupidest idea I have ever heard. What do you think? Your bias affects how people answer. Revised: What is your opinion of Via Verde?
People Like to Help You They might say what they think you want them to say! That is why you must avoid bias. ALSO: Avoid Leading Questions AND: Always give an option maybe not sure don’t know What’s the difference between “I don’t know” and “I don’t care” ??
People might answer automatically LEADING QUESTIONS You don’t like Via Verde, do you? Via Verde is a good idea, isn’t it? Do you think it’s a good idea? Do you think it’s a bad idea? NEUTRAL: What is your opinion of Via Verde?
Avoid Using Negatives Do not use the word NOT! It is too confusing. BAD QUESTION: The food in the cafeteria is not good. yes no Yes, it’s not good? Or No, it’s not good? Or yes it is good, or no, it is not not good?
Negatives only complicate things “I think it is good.” means the same as “I do not think it is not good.”
Yes/No Questions You can use these but you will need more information from your participants. So, ask them yes/no and then also ask them some other kind(s) of question(s) to know why they answered yes or no.
Remember… You need enough information to write about and answer your research question. But Do not take information you don’t need
Make the question reasonable How much TV do you watch every week? 15 minutes 30 minutes 60 minutes 75 minutes More than 75 minutes
Be sure your choices make people feel normal How much TV do you watch each week? Less than 1 hour 2 hours 3 hours 4 hours or more This makes the 4 hour person feel they are at the end of the scale – they might think it is better to answer 3 hours. But if “4 hours” is in the middle, then they will not feel odd at all.
You will get more meaningful results How much TV do you watch each week? 1 hour or less 2 hours 3 hours 4 hours 5 hours 6 hours 7 hours 8 hours or more
Or use an open-end number How many hours of TV do you watch each week? ________ hours (approximately)
More unreasonable questions How often do you text while driving? times per day More than 100 times per day
Likert Scale of Agreement Usually, use an odd number (3 or 5) so you have a middle Very some neither some very Much what agree what much Agree agree nor disagree disagree disagree
Likert Scale gives you a number Yes/No gives you a percentage. Just over half, or 55%, of the participants agreed. Likert Scale gives you a number based on the scale. The mean (“average”) response was 3.2, or just slightly above “no opinion.”
Rating Scale I think the idea is A.Excellent B.Good C.Fair D.Poor E.No opinion
Using an Even Number Some researchers like to force the answer. They do not give a middle “no opinion.” Strongly some some strongly Agree what what disagree agree disagree (You have to decide if that is what you want to do – some people will refuse to answer or they will get upset) Always put AGREE first and DISAGREE last. This is the order people expect.
Giving the Option to Not Answer Many times, people will not answer if they do not see an answer they like. Other times, people will answer even though they don’t like the answers. Then they will be irritated and you will get the wrong result. THEREFORE, ALWAYS GIVE AN OPTION: don’t know maybe not sure no opinion need more information I don’t care N/A
Multiple Choice: It’s Not Easy! Multiple Choice is good when you are looking for something specific (this, this, or this). BUT If you want to know what people really think, take time to make really good questions. If you use multiple choice, say “check all that apply” and also have “NONE” and “OTHER” – but usually people will not write anything after they check that. So you will not have any information.
That’s Why It’s Called… MULTIPLE CHOICE People have to choose from the choices you give. So that is good if that is what you want. BUT do not make the list too long. The person might start having to think a lot and then be unable to decide. If you want true opinions, limit your use of multiple choice. Use them combined with short answer. Ask people “WHY” in a separate question.
SUMMARY: Two kinds of questions 1.CLOSED END: Yes/No Likert Scale/Rating Scale Multiple Choice 2.OPEN END: 1.Words I do not recycle because ____________________________________________ ___________________________________________. 2.Number I recycle ______ times per month.
How Do I Organize My Questions? Put the easy / happy ones first Group questions on the same idea together Put difficult or controversial questions at the end If you put the hard ones last, the person will feel better and be worried for less time. If you put the hard ones first, they might not answer the rest of it so well. Of course you do not want to upset anybody, but it can happen with sensitive or political topics.
If I have a combination of question types… It is good to mix them up. You can put some yes/no and then have some Likert scale. Then have an open ended written answer. Then another yes/no and some multiple choice. Doing this is good because it makes people read the questions. If they see 12 questions all the same format, they might start to answer automatically without really thinking about it.
Making your Questionnaire 1.Write a draft of your questions. Be sure you have enough questions to find out some useful information and details. 2.You must test your questions on at least three people during the class – get their reactions. 3.Do not give your questionnaire to anybody else until the professor has seen it and signed her approval.
Finally … Think about how you feel when you answer a questionnaire. What do you like or not like about it? What makes it friendly or nice to do? What makes you feel they appreciate your help? How do you like it to look on the page? What kinds of questions are good, and what kinds make you feel you cannot tell the real answer? (do you like to take multiple choice tests?)