Chapter 7 Survey Research
3 additional guidelines: Minimizing Fence Sitters and Floating
Fence Sitters: People who see themselves as being neutral, may skew the results if you force them to choose between opposites; example Floaters: Respondents who choose a substantive answer when they really don't know the answer; example: a third of public will provide a response on law that they know nothing about if there is not a "don't know" option
Three features of survey research: Generalizability
Most effective way to obtain information from representative sample of large population; how do we really access a representative sample?
What does pretesting the survey consist of?
Start by discussing the content with other researchers (may bring new ideas to your project). Type of survey depends on a) complexity of questions b) resources c) characteristics of respondents d) type of survey design
What do you want to do in a questionnaire?
Start with a problem --> research questions --> hypothesis identify --> questionnaire; Maintain a consistent focus (include only relevant questions) Build on existing instruments (can improve reliability and validity, use already designed ones)
What is Mixed Mode Survey?
Surveys that are conducted by more than one method Allow strength of one design to compensate for weakness of another Modes selected depend upon population being surveyed
What are face to face surveys?
o Highest response rate o Most expensive, especially if geographically spread out o Can be easy to make up responses, especially in field interviews Interviewers can make observations about interview setting or environment
What are telephone surveys?**
o Highest response rate for lowest cost o Interviewers can be more closely supervised than in in-person interviews o Safer for interviewers??
What are the 5 basic types of surveys?
1) mailed 2) group administered 3) telephone 4) In person 5) electronic
What are the advantages and disadvantages of open ended questions?
- Detailed responses - No preconceived content of answer - Social desirability: what is that you want me to say - Must go through several steps before ready for analysis
Explain more about Fence Sitters:
- To find out why they are neutral provide a neutral or undecided response option - Drawback: may encourage respondents to take the easy way out and choose neutral or undecided just because
What are three features of survey research?
1) versatility 2) efficiency 3) generalizability
Why are there issues with validity in telephone surveys?
1. Not reaching proper sampling unit - Random digit dialing (RDD): Machine calls phone numbers randomly within designated area codes and exchanges, regardless of whether the numbers are published 2. Not getting enough responses to have generalizable results - Use Computer Assisted Telephone Interviewing (CATI) ♣ Data collection and data entry occur concurrently ♣ Several machine edit features help minimize date entry error - Use Computerized Interactive Voice Response (IVR) ♣ Interviewers receive automatic calls and answer by using keypad Record voice responses that are transcribed later
What is a scale?
A composite measure of one concept created from a series of two or more questions
What is a questionnaire?
A self-administered survey
What is a reference period?
A time frame in which a survey asks people to place a certain time period (in the last 6 months)
What are two components of avoiding confusing phrasing?
A) Double negatives Question containing two negatives making the question unclear "Do you disagree that there should not be a death penalty?" B) Double barreled Question about more than one thing "Do you know anyone who has used cocaine?" ___ I have used cocaine ___ I have not used cocaine
5 basic types of surveys: electronic
A) E-mailed - sent as message to respondent's email B) Web Based: useful for growth in internet and technology makes it easier ADVANTAGES: ability to control question sequence, can ask complex questions, low cost DISADVANTAGES: Large portions of households without internet access
What are the 3 additional guidelines for Fixed Response Questions
A) Likert Questions B) Minimizing Fence Sitters and Floating C) Filter Questions and Skip Pattern
What are the 5 basic principles of writing survey questions?
A) Writing Clear and Meaningful Questions B) Avoid Confusing Phrasing C) Minimize the Risk of Bias D) Avoid Making Either Disagreement or Agreement Disagreeable E) Maximize the Utility of Response Categories
What are the advantages and disadvantages of closed ended question?
A: Quick response (sometimes you don't want the ENTIRE story) Consistency in type of response Easier to analyze statistically
What are demographic questions?
Basic description of members of a population; must be relevant to your study (age, sex, gender, employment status)
What is a open ended question?
Allow respondent to fill in own answer May ask for specific information or May ask for narrative response Respondent writes whatever he/she wants Also known as interpretive question
5 basic principles of writing survey questions: Minimize Risk of Bias
Biased or "loaded" words and phrases tend to produce misleading answers; no valid data if you pressure people to feel they cannot be honest
What is split ballot design?
Certain people are given a survey with certain questions and others given another set of questions Reason: determine the effectiveness of questions
5 basic types of surveys: Group administered
Completed by individuals assembled in a group Response rate is not usually a concern Drawback: assembling a group, respondents may feel coerced to participate and will not answer honestly, and feel the researcher is using another organization
5 basic types of surveys: What is a mailed survey?
Conducted by mailing a questionnaire to respondents ADVANTAGES: Ability to control question sequence, can implement quickly DISADVANTAGES: difficult maximizing the response rate People may stop answering questions or simply not reply
Explain self reports and criminal activity?
Criminal activity determines 1 or both of: - Prevalence (proportion of sample engaging in "") -Incidence (how frequently)
Three features of survey research: Efficiency
Data can be collected from a large number of people relatively inexpensively and quickly
3 additional guidelines: Filter Questions and Skip Pattern
Filter question: Question asked to establish whether concept is applicable Skip pattern: Questions skipped by respondents to whom they are not applicable Contingent questions: Question or set of questions answered depending on the Q before it
What are close ended questions also called?
Fixed choice questions Forced choice questions
What is structure?
Fixing in advance the content and order of questions and response choices
What is a reverse code?
Have you eaten breakfast this morning? Yes Did you skip breakfast this morning? No These questions are asking the same question and should yield the same response. This is how you would measure if participants are just answering ignorantly or not really paying attention to the questions
What is survey Research?
Info collection from responses to questions;
What is IVR?
Interactive Voice Response (IVR): Respondent receives automated phone call from 'computer' and answers questions by pressing numbers on phone keypad or speaking answers
5 basic types of surveys: Telephone
Interviewers question respondents over the phone and record answers ADVANTAGES: Ability to include complex questions and open questions, ability to control question sequence DISADVANTAGES: Issues with Validity
What is the Omnibus Survey?
Large-scale study that covers a range of topics in single survey
In what ways can self-administered surveys be given?
Mailed In Person Electronically People fill in own answers Most economical way to survey large number of people
What is anonymity?
No identifying information is recorded that could link back to you by the research
What are the two basic types of questions?
Open ended and Close ended
What is the response rate?
Percentage of persons surveyed who complete a survey Item response rate - percentage of items that are completed on a survey [There are questions that respondents will not answer - gaps]
What are the ethical issues in survey research?
Protection of Respondents o Potential harm must be disclosed in cover letter o Make every effort to reduce emotional trauma affiliated with survey participation Confidentiality o Socially stigmatized or illegal behaviors must be kept in strictest confidence o Use ID numbers for respondents
How do you avoid floaters?
Provide detailed information about the issue when asking the question
Three features of survey research: Versatility
Researchers can ask about almost any topic ; Enhance understanding of social issues; Can be designed to study almost any social issue; Can learn about individuals and organizations
What is a closed ended question?
Respondent chooses from list of possible responses
What are some guidelines for questionnaires?
a) Major topic divisions b) Instructions c) Attractiveness d) Response choices should be designated by numbers to help with data entry
What is an interview schedule?
Survey instrument containing the questions asked by the interviewer in an in-person or phone survey
What are interpretive questions?
Survey questions used to elicit additional information following an open-ended question
5 basic principles of writing survey questions: What is writing clear and meaningful questions?
Surveys are not conversations Survey questions must asked of many people and used the same for each person Everyone should understand the survey the same way
What is a cover letter?
The letter sent with a mailed questionnaire. It explains the purpose encourages people to participate
5 basic principles of writing survey questions: Maximize the Utility of Response Categories
The response choices must be exhaustive [make sure all categories are accounted for] and mutually exclusive [categories that can be separated out] Example: How many times have you been arrested? 0 1-2 3-5 6-10 11 or more (No overlapping)
5 basic types of surveys: In person (aka Interview schedule)
Types ¥ Manual or Computer Assisted Personal Interview (CAPI) ¥ Computer-Assisted self-interviewing (CASI) Have better quality data than mail surveys ¥ Interviewer can clarify unclear items ¥ Interviewer can "probe" for open-ended responses ¥ Usually requires extensive training ¥ High response rate ADVANTAGES: Can avoid social desirability bias, contamination, high representative sample DISADVANTAGES: Time-consuming and costly
5 basic principles of writing survey questions: Avoid Confusing Phrasing
Use correct grammar Use shorter rather than longer words and sentences Break complex ideas into several questions and address each component separately
3 additional guidelines: Likert Questions
Usually framed as declarative statements - not questions React to statement by identifying intensity of feeling/belief/opinion (1 though 4) (Can be with or without a neutral point)
What is idiosyncratic variation?
Variation in responses to a question that is due to reactions to particular words or ideas in the questions
What is an example of a close ended question?
What is your annual income? Check the appropriate box [Ordinal LOM]: ▫ No income Less than $5000
What is an example of open ended question?
What is your date of birth? ____/_____/_____ Describe the most difficult thing about being a single, working parent. _________________
What are Filter Questions and Skip Patter used for?
When a question or group of questions applies to only some - reduces total time to complete - Reduces respondent burden (I am tired of answering these questions)
When do you combine questions into an index?
When several questions measure one concept; Sum of responses = a composite measure which tells to what degree the concept is present Example: numerous questions on deviant behavior y=1 no=0. The more yes answered questions the more deviant the person
5 basic principles of writing survey questions: Avoid Making Either Disagreement or Agreement Disagreeable
Which question are people more likely to answer honestly? "Have you ever taken anything from a store without paying for it?" [Yes I have ever taken... or No I have never = this is confusing] OR "Have you ever shoplifted something from a store?" [Better question]
What are some of the things surveys are used for in CJ?
a) Counting crime b) self reports c) perceptions
What are some of the considerations in creating indexes?**
¥ Assumption that each question measures same concept may be mistaken ¥ Combining responses to specific questions can obscure important differences in meaning ¥ Questions may cluster together in subsets?? ¥ Sometimes particular questions are counted or weighted more than others in calculation of the index
How should organization matter?
¥ Create a title that is descriptive of overall topic ¥ Make sure question order does not influence responses ¥ Create major topic divisions to help respondents understand organization of questionnaire ¥ Use instructions throughout questionnaire to minimize respondent confusion