Research Methods Unit 3 Survey Research research o Most widely used method technique for data gathering for social o We can ask people about their behavior Have you done this Don t want to be too ambiguous o We can ask about attitudes beliefs opinions o We can ask people about their characteristics Age marital status ect o Self classification Asking people to put themselves into a category that may not be straight forward Example race Liberal vs conservative o We can ask people about their expectations Do you plan to move to a new home in the next year Do you plan on buying a new car in 6 months In criminology we ask people about hypothetical crime o We can ask people about knowledge Sometimes we want to gauge how much people know You can ask Who is our current mayor o 2 forms of questions Open ended Have not answer choices provided free response Advantages o Allow us to have more in depth analysis Disadvantages o A lot of work o Must use something called coding You read through each answer and place them into categories then each category is given a number Relies on the interpretation of the researcher o Some answers given can be completely Close ended irrelevant Have response choices provided Much more common Advantages o Easily coded Disadvantages o Have to be mutually exclusive and exhausted o Rules for question wording An option for everyone and only one option for everyone In general use 8th grade reading level for vocabulary 1 Avoid jargon slang and abbreviation Usually don t want to use misdemeanor or felony Bad example who will win this weekend FSU or USC Example don t use slang names for drugs 2 Avoid ambiguity confusion and vagueness What is your income 3 Do not use indefinite words or response categories 4 Avoid emotional language Example welfare vs government assistance o Welfare is much more emotional Try to be as neutral as possible 5 Avoid prestige bias Example 4 out of 5 doctors believe this Just ask the question without associating it to someone 6 Avoid double barreled questions Example do you think the U S government should spend more money on welfare or national security Break down that question into 2 questions 7 Avoid asking someone to identify the relationship Example do you like politicians that are married better than those that aren t 8 Avoid leading question These questions lead respondents to answer one way Example ohhh you don t smoke do you It leads the respondent to chose one response over the other do to wording 9 Avoid loaded questions Example should the mayor spend more money on cleaning up the streets vs should they mayor spend to fix the potholes on the dangerous streets Different from emotional language because you are just using langue to stimulate a positive or negative response not evoke an emotional reaction 10 Avoid questions that are beyond respondents capabilities Example where were you three years ago yesterday Example who is the prime minister in Canada Can be because of a memory issue or because its something they truly don t know Its bad information because people will answer anything It can also be discouraging because people really need to think about it 11 Avoid false premises Anytime we make a statement from the start that the respondent may not agree with Example the library isn t open enough Would you prefer it to be open two hours earlier or stay open two hours later 12 Avoid double negatives Example do you agree or disagree that students should not be required to take an aptitude test to get into college 13 Avoid unbalanced response categories Example how well do you think your teacher is doing o Answers excellent very good or good There is no bad answer 14 Keep questions short Respondents shouldn t have to study the question you have a little more leeway if it is a written test survey but if it is over the phone then keep it short if its too long it may be skipped or if they do answer their stamina will be reduced respondent fatigue People do this because they want their questionnaire to feel short 2 Contingency questions Really common in research I ask you a question and depending on your answer you either continue to the next question or may skip ahead to a different question You want to be as clear as possible 3 Matrix questions Really useful when all you re questions are of the same Kind of like a graph what you see on most surveys Allows you to get more questions in a smaller amount of format space We want to mix up positive and negative worded questions Sometimes as a researcher we want to have our questions to fit into a matrix so we may not word questions as good as they should be worded Questionnaire construction o How do we perfect our questionnaire 1 Don t try to cram too much on one page 4 Order for questions The order the questions are in actually affects how respondents answer the question Some researchers put questions in a totally random order o This is not efficient for the survey Its best to put the most interesting questions first o Put demographic questions last o Exactly the opposite when it is an in person interview Start with the super boring questions Pretesting actual survey o Going out and administering your survey before you conduct the o Important because it tests your survey to make sure it is efficient o Helps you to catch mistakes o Ideally your sample for the pretest should be similar to that of the real survey Response rates o What is a considered an acceptable response rate Depends on the survey 70 is usually shockingly good 50 is still considered pretty good Some research indicates that 30 40 is still acceptable o We like to have some information about people who don t take our survey Types of surveys o Self administered So we can see if they are similar in some way You are reading the questions yourself Kinds Mailed surveys Internet surveys Written questionnaires Advantages Really cheap You can usually administer these by yourself Really easy to keep anonymity Not anonymous if it is a longitudinal study that I will need to know your information specifically Avoid any potential interviewer bias You are reading your own questions and responding with your own answers An actually give you really high response rates if you reach a population that is well educated or interested in the topic Disadvantages Could potentially take a long time o Most return in the first 2 weeks o But it could take months to get a response Virtually no controls over how the survey is taken o The survey may be taken
View Full Document
Unlocking...