COMSTRAT 309 Sampling Notes
COMSTRAT 309 Sampling Notes 309
Popular in Comstrat- Quantitative Research Methods
Popular in Communication
This 6 page Class Notes was uploaded by Alison Wuerch on Thursday March 5, 2015. The Class Notes belongs to 309 at Washington State University taught by Hmielowski in Spring2015. Since its upload, it has received 70 views. For similar materials see Comstrat- Quantitative Research Methods in Communication at Washington State University.
Reviews for COMSTRAT 309 Sampling Notes
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 03/05/15
COMSTRAT 309 Qualitative research methods Week 3335 1 Descriptive and Analytical Surveys a There are two types of surveys i Descriptive 1 Attempts to describe or document current conditions or attitudes ii Analytical 1 Attempts to describe and explain WHY situations exist a Two or more variables are usually examined during this approach i Research questions ii Research Hypothesis Sampling error example a Size doesn39t always matter i Which is better 1 Sample of 24 million or sample of 50k ii Error matters Error a Reasons for the Research error has two types i Sampling error too man y 1 Democrats republicans etc a Sample doesn39t match reality ii Non sampling error measurement error 1 Random error Survey questions gives you different results across samples assumes sample is correct a Very hard to detect i Poor precision ii Nonspeci c causes iii Not reproducible 2 Systematic error Results wrong but always in the same direction a Can gure out and correct the problem i Poor accuracy ii De nite causes iii Reproducible Problem of NonResponse a You can randomly select people but you cannot make people participate b Nonresponse destroys the generazibility of the sample i You are generalizing to people who are willing to respond to surveys NonProbability Sampling a Any method where a member of a population does not have an equal change at being selected b Most cases these cannot be generalized c Always asking for speci c people 6 Convenience Sampling a When subjects are selected based on availability to the researcher b Selection is generally NOT tied to purpose of the research 7 Purposive sampling a Select certain individuals for special reasons b Subjects selected for a good reason tied to the purposes of research c Useful for populations that are not easily obtained by screening general population 8 Volunteer sampling a Individuals volunteer to be included i Extra credit ii Try out a product before its released iii Discounts 9 Quota sampling a Research samples with a certain number of subjects in various categories i 100 males and 100 females 10Snowball sampling a The researcher approaches on subject asks that subject to suggest to others and it continues on 11Questionnaire design a Survey research i Anytime you systematically ask people about their attitudes emotions beliefs knowledge intentions or behaviors 1 Professor evaluations 2 Public opinion survey b Every question in any questionnaire even the simplest question should include some form of quotDon39t Know No Answerquot 12Primary Goals a Identifydescribe attitudes or behaviors in a given population b Examine relationships between variable measured i Does x relate to y 13Types of survey questions a Open ended i Participants generate responses 1 Ex What could your favorite radio station change so that you would listen more often ii Advantages 1 More detail and in depth responses 2 Allows for unforeseeable responses a Answers that may suggest possible relationships with other answers or variables 3 Allows freedom to answer 4 Very useful in pilot tests in a study iii Disadvantages 1 Very time intensive a A lot of time needed to collect and analyze all the responses 2 Can be hard to analyze Coding a Because there are so many types of responses a content analysis of each open ended question must be completed to produce data that can be tabulated b Close ended i Participants choose from provided responses ii Advantages 1 Uniform responses allow for easy comparability 2 Quick analysis 3 Easy to quantify iii Disadvantages 1 May have a different answer from those that are supplied a May force respondents to choose a response that they might not completely agree with b Easy way to x this is to provide an answer of quotOtherquot 14Close ended Questions a Check all that apply b Multiple choice c Semantic differential scales d Likert scales e Rank ordering 15Question Guidelines a Be Clear i What might be perfectly clear to the researchers might not be nearly as clear for the respondents ii Avoid dif cult or specialized words acronyms or pretentious language iii Use the vocabulary used in Newspapers or popular magazines b Keep questions short i Many respondents are in a hurry ii They will not take time out of their busy schedules to take your survey unless its short c Avoid negativesdouble negatives d Avoid double barreled questions i 2 in 1 questions ii Whenever the word and appears in a question the sentence structure should be examined to see whether more than one question is being asked e Avoid leadingbiased questions iEx quotIn your free time would you rather read a book or just watch televisionquot LThe word justin this example injects a bias into the question Lit implies that there is something less desirable about watching television ii A leading question suggests a certain response or contains a hidden premise f Avoid asking about issues beyond participants capabilities of Mo iyou want everyone to be able to answer your question to the best of their ability iiThis sample should all have an in depth answer on the question 0 Avoid false premises h Avoid asking embarrassing questions unless necessarv i Put them at the end ii Begin with the easy questions 1 Questions about income may cause the respondent to feel embarrassed or pressured to answer a different answer than they would usually give i Resoonse Options should i Match the question 1 15 scale ii Be mutually exclusive 1 Can only select 1 2 No overlap a Don t want any confusion iii Be exhaustive 1 All possible options are available 2 Needs a range of options iv Equal number of positives to negatives in response answer 16Phrasing Concerns a Social desirability bias i Over reporting desirable behaviors or underreporting undesirable behaviors 1 There is a social pressure to make ourselves look better and feel good b Two ways to reduce desirability bias i Emphasize anonymity 1 Every person must be anonymous ii Phase questions in a way to 39safe face 1 Always over representing ourselves in good situations 2 But underrepresenting ourselves in bad situations a Drinking b Drugs 17Questionnaire Design a Introduction one of the best ways to increase the response rate i Need to introduce yourself ii What you are doing with the survey iii And why what will it be used for 1 Talk about the risks and bene ts required 2 Talk about the time needed iv Emphasize that you are NOT selling anything v Selfexplanatory 1 Need clear questions 2 If you are sending it over mail there is no way for them to ask about a question vi Only ask necessary questions 1 There is a limited time frame a Usually 1520 minutes vii Use a limited number of question styles 1 Don39t use many different scales 2 If you use different scales try and clump them together a Don t break them up viii Make sure it is easy to read and follow 1 Skip patterns a If you answered no skip to question 3 if your answer is yes proceed to question 2 ix Warm up participants 1 Have the easy questions at the beginning and the sensitive ones last x Order effects 1 Is the question asked going to relate at all to the next question asked b Always Consider Will questions that you are asking in uence questions that follow 18Question Order a All surveys ow better when the early questions are simple and easy to answer b Usually start with a quotwarm up icebreakerquot i Want the respondents to become more accustomed to answering questions with other people c Length i The length of the questionnaire is directly related with the completion rate 1 Long questionnaires cause fatigue respondent dropout and low completion rates 2 Shorter questionnaires guarantee higher completion rates ii There is no strict guidelines on the length and will depend on a variety of factors
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'