Com 309 Exam 2 WSU fully solved and updated

What is a control variable? - answer-Used to ensure the results of the study are due to the independent variable, not to another source (used to eliminate unwanted influences) What is noise? - answer-all the variables that may create misleading results (ex. abstract answers to open ended questions, people unknowingly providing false information) Measurement - answer-When a researcher assigns numerals to objects, events, or properties according to certain rules Numerical (no implicit quantitative meaning), assignment (designation of numbers to certain objects or events), and rules (specify the way that numerals or numbers are to be assigned) - answer-What are the three components of measurement? Rules - answer-Which component of measurement is this: specifies the way that numerals or numbers are to be assigned ex. measuring speed Assignment - answer-Which component of measurement is this: designation of numbers to certain objects or events (ex. designating the numeral 1 to the people who obtain most their news from tv, numeral 2 to people who get it from newspapers, etc) Numerical - answer-Which component of measurement is this: has no implicit quantitative meaning. When it is given meaning it becomes a number Dummy variable - answer-Numerals that are used as labels to represent categories, but have no mathematical significance Ex: Republican: 0 Democrat: 1 Independent: 3 Other: 4 (this could be interpreted that independents are 3 better than republicans) 1. A scale with more points rather than fewer points allows for greater differentiation on the item being rated 2. People of all age groups, race, and nationalities like to use a 1-10 scale. (it's universal, everyone understands) 3. When using simple rating scales, it's best to tell respondents "The higher the number, the more you agree" or "the higher the number the more you like" - answer-What are the 3 things to consider when selecting the type of scale you want to use in your study? Scale - answer-Def: a composite measure of a variable; it is based on more than one item 1. Purpose of the study (is it designed to generalize results or something else?) 2. Cost versus value (probability sampling costs more) 3. Time constraints (probability sampling is more time consuming) 4. Amount of acceptable error (when error control is not a primary concern, use non-probability) - answer-What are the four things to consider when making a decision about whether to probability or non-probability sample? Unqualified volunteer sample - answer-What kind of sample is this: Researchers have no control over the respondents or subjects who participate in a study--virtually anyone can participate Qualified volunteer sample - answer-What kind of sample is this: Selected using probability sampling and employs controls to limit the type of person who is invited to participate in the project Sampling frame - answer-a list of the items or people forming a population from which a sample is taken 1. project type 2. project purpose 3. project complexity 4. amount of error tolerated 5. time constraints 6. financial constraints 7. previous research in the area - answer-What are the seven factors that determine the sample size required for a study? Weighting/sample balancing - answer-When the subject totals in given categories do not reach the necessary population percentages, subjects' responses are multiplied to allow for the shortfall (highly controversial) a survey that attempts to describe or document current conditions or attitudes (to explain what exists at the moment) - answer-What is a descriptive survey? a survey that attempts describe and explain why situations exists - answer-What is an analytical survey? Descriptive (used to gather info on what people do and think) and analytical (used to answer research questions or test a hypothesis) - answer-What are the two major types of surveys as described in the book? 1. They can be used to investigate problems in realistic settings (ex. newspaper readings, television viewing, consumer behavior patters can be examined where they happen rather than in a lab) 2. The cost of surveys is reasonable for the amount of info gathered 3. A large amount of data can be collected with relative ease from a variety of people 4. Surveys are not constrained by geographic boundaries 5. Data that are helpful to survey researchers already exist - answer-What are the five advantages of survey methods? 1. Independent variables cannot be manipulated the way they are in lab experiments 2. Inappropriate wording or placement of questions can bias results 3. The wrong respondents may be included in survey research 4. Response rates are declining - answer-What are the four disadvantages of survey methods? a measurement of the amount of people who respond to a survey - answer-What is a response rate? Used to eliminate unwanted respondents and include only respondents who have specific characteristics --A short survey question or a portion of a survey designed to select only appropriate respondents for a research project - answer-What is a screener/filter question? 10% - answer-What would be the number of 'breakoffs' that would suggest the questionnaire is too long? (ex. someone hangs up in the middle of an interview) offer.... 1. 2. Prepaid phone calling cards activated when the questionnaire is returned 3. gift cards to local retailers 4. cash (no less than $10) - answer-What are the four ways to increase response rates in mail surveys as much as 50%? 1. Subjects are often unable to recall information about themselves or their activities 2. People provide false answers so they don't have to admit that they don't know 3. Subjects may purposely deceive researchers by giving incorrect answers to questions 4. Respondents give elaborate answers to simple questions to try to figure out the purpose of the study 5. People are unable to put their true feelings, perceptions, and beliefs into words - answer-What are the 5 general problems associated with survey research? conceptual (constructive) definition - answer-dictionary definition using words to define operational definition - answer-def: procedures to measure a concept goal operationalization - answer-def: the process of creating an operational definition is called __________ top: conceptual bottom: operational - answer-EXAMPLE CONCEPTUAL/OPERATIONAL: Athleticism: "physically active and strong, good at sports" vs Measuring by balance, flexibility, running a mile NO - answer-Is this a good fit between operational and conceptual definitions? conceptual: violence operational: measure the # of times people are shown hugging, helping, being kind etc. consistency across groups - answer-How do you know if you have a good measure? validity - answer-def: When you are actually measuring what you think you're measuring reliability - answer-def: measuring a concept consistently (provides consistent results) validity - answer-Bulls eye metaphor: hitting the target represents ____ reliability - answer-Bulls eye metaphor: hitting in the same spot represents ______ 1. Face validity 2. Content validity 3. Criterion/predictive validity 4. Concurrrent validity 5. Constructive validity - answer-What are the 5 types of internal validity? Face validity - answer-What type of internal validity is this? the measure seems to look good on the face of it (looking at it, it looks right) - should be done by experts not tied to the study ex. i tend to get in a lot of arguments compared with how many drinks do you have a week Content validity - answer-What type of internal validity is this? The concept measures the full range/all aspects of meanings/dimensions of the concept ex. measuring depression should measure psychological and behavioral Criterion/predictive validity - answer-What type of internal validity is this? The measure is shown to predict scores on some other relevant feature measure ex. 10th grade IQ measuring college GPA Concurrent validity - answer-What type of internal validity is this? Measure discriminates between groups that should differ on key variable ex. tests for athleticism separate people who work out from those who don't Construct - answer-What type of internal validity is this? The measure is shown to be related to other concepts that it should be related to (and unrelated to ones that it shouldn't be related to) ex. Grit should be related to effort, grades, determination, should not be related to anger, shyness stability - answer-Testing people the first week of the semester and the last week of the semester is a way to test __________ Internal consistency *has to do with reliability - answer-examining the consistency of performance on items that compose a scale Split-half technique - answer-Ways to test internal consistency: A technique related to internal consistency where the score on the first half of the test should correlate highly with the score on the second half Cronbach's Alpha - answer-Ways to test internal consistency: assess internal consistency of a measure by assessing whether items get at the same concept Equivalency cross test: * *has to do with reliability - answer-Ways to test reliability: assesses the relative correlation between two parallel forms of a test Intercoder reliability *has to do with reliability - answer-Ways to test equivilency: Seeing if results can be reproduced by others (used most often in content analysis) Independent Variable (IV) - answer-The variable that is the "Cause" or the "Predictor" Dependent Variable (DV) - answer-The variable that is the "effect" or the "outcome" Nominal - answer-Levels of measurement: - Variable is measured with categories - Numbers here are meaningless - Categories must be mutually exclusive and exhaustive -ex. What political party do you identify with? Ordinal - answer-Levels of measurement: - Measured with rank-order category - Tells you placement, but not how far apart - ex. ranking of sports teams: you can tell what order they're in, but not how far apart they are Interval - answer-Levels of measurement: - measurement points with equal distance between them - no true 0 ex. temperature has no true 0, and it takes the same amount of heat to go from 30 to 40 degrees as it does form 40 to 50 degrees Ratio - answer-Levels of measurement: - measurement points with equal distance between them - there is a true, meaningful 0, 0 is important for stats - ex. Time and distance (you can say a car going 50 mph is going twice as fast as 25 mph) discrete - answer-Discrete vs continuous: - nominal - measure doesn't include subparts - ex. number of children in a family Continuous - answer-Discrete vs continuous: - interval, ratio - can have sub-parts, ex 32.5 degrees, or height Likert Scale - answer-A scale where participants have the option to strongly agree, agree, be neutral, disagree, or strongly disagree with a statement good fit between your operational and conceptual defs - answer-Conceptual fit: Stability - answer-The consistency of a result or of a measure at different points i time Nominal - answer-Is this nominal, ordinal, ratio or interval? Which political party do you identify with? 1. Republican 2. Democrat 3. Independent 4. Green 1. Mutually exclusive (no overlap in the categories) 2. Exhaustive (everything must fit into a category) - answer-What two things are needed to have an appropriate nominal measure? NO - answer-Do numbers mean anything for a nominal measure? Ordinal (because we can tell what order they're in but not how far apart they are) - answer-Is this nominal, ordinal, ratio or interval? Rank your major news sources 1. Internet 2. Tv 3. Radio 4. Newspaper Ordinal (because we can tell what order they're in but not how far apart they are) - answer-Is this nominal, ordinal, ratio or interval? "How much TV did you watch yesterday?" - a little - a lot - none Interval (because there are measurement points with equal distance between them but there is no true 0) - you could not say that 20 degrees is twice as hot as 10 degrees because 0 degrees does not mean there is no temperature - answer-Is this nominal, ordinal, ratio or interval? The temperature (there is no such thing as no temperature) Ratio (because there is measurement points with equal distance between them and there is a true 0 that is important) - answer-Is this nominal, ordinal, ratio or interval? Time and distance (you can say a car going 50 is going twice as fast as a car going 25) Semantic Differential - answer-Participants are given two opposite words on opposing sides of the scale Motivated 1 2 3 4 5 6 7 Unmotivated Uninterested 1 2 3 4 5 6 7 Interested Involved 1 2 3 4 5 6 7 Uninvolved Isomorphism - answer-Whether a measure corresponds with reality (is a 2 on an attitude scale actually a 2) Sampling - answer-Def: describes how you select your subjects Representative - answer-This kind of sampling can be used to generalize to a larger population Non-representative/non-probability sampling - answer-This kind of sampling is used when you are looking for a specific characteristic - can't be generalized ex. internet polls Population ex. Population: all WSU students, if we were able to actually survey all of them it would be a census - answer-The entire group of people who you want to study ex. All people over 18 living in the US ex. All registered voters who are likely to vote Census ex. Population: all WSU students, if we were able to actually survey all of them it would be a census - answer-When the entire population is studied - very rare - very expensive and time consuming 1. Obtain a list of all population members 2. Assign #'s to all members 3. Randomly select #'s until desired sample is reached ex. could get list of all WSU students ID numbers and randomly contact by using random number generator ex. Phone surveys; random digit dialing - answer-Describe how to create a simple random sample 1. Obtain list of all population members 2. Assign numbers to all people (members) 3. Randomly select start position on the list 4. Select every 'Kth" element from the list ex. Divide # of population (1000) members by desired sample size (100) so you select every 10th person....start with 5,15,25,35 - answer-Describe how to create a systematic random sample Systematic random sample Randomly selecting 4 digit numbers and adding them to the three digit exchange prefixes in the city where the survey is being conducted - answer-What type of sampling is being used for random digit dialing? When you don't have a list of people 1. randomly select clusters (ex. Zip codes) 2. randomly select participants from within the clusters - answer-Why would you use multi-stage cluster sampling? How do you do it? To get representation from a certain group (Strata are variables/characteristics that are important to you) 1. Divide population into strata (groups) 2. Then you randomly sample from each group until you reach desired sample ex. Could be class level @ WSU and you want 35% freshman, 30% sophomores, 20% juniors, 15% seniors - answer-Why would you use a stratified random sample? How do you do it? Margin of error - answer-The plus or minus percentage range of how close your results are to the actual population Confidence Level - answer-The certainty that your results fall within a certain range Sampling error - answer-The degree to which measurements obtained from a sample differ from the measurements that would be obtained from the population Random error (very hard to detect the source of error) - answer-Non-sampling error: Survey question give different results across samples Systematic error (might be able to figure out and correct) - answer-Non-sampling error: Results are wrong, but always in the same way Non-response Non-response is a problem because you can only generalize your findings to people who respond, and there is some evidence to suggest that people who are willing to fill out survey answer differently than those who aren't willing - answer-You can randomly select people but you can't make them answer. This is called ____________ Why is this a problem? Non-probability sampling (usually can't be generalized) - answer-Any sample where a member of a population doesn't have an equal chance of being selected 1. Simple Random Sample 2. Systematic Random Sample 3. Stratified Random Sample 4. Multistage Cluster Sample - answer-What are the 4 types of probability samples? 1. Convenience 2. Quota 3. Snowball 4. Volunteer 5. Purposive - answer-What are the 5 kinds of non-probability sampling? Convenience - answer-What kind of non-probability sampling does this describe? When subjects are chosen based on availability to the researcher, not the purpose of the project ex. Wanting data from students so you go to a school close to you Volunteer - answer-What kind of non-probability sampling does this describe? Individuals volunteer to be included in a study ex. extra credit, try new products, discounts Purposive - answer-What kind of non-probability sampling does this describe? Select certain individuals for specific reasons tied to research project ex. Meth users, victims of abuse Quota - answer-What kind of non-probability sampling does this describe? research sampling with a certain number of subjects in various categories, but does not use random sampling method ex. Try to get samples to look like population (i.e 50% males 50% females) *NON-PROBABILITY VERSION OF STRATIFIED RANDOM* Snowball - answer-What kind of non-probability sampling does this describe? The researcher approaches one subject, asks that subject to suggest others, they suggest others, and so on (useful for hard to reach subjects) ex. Homeless people, Bette Midler 1. Identify attitudes or behaviors 2. Examine relationships between variables measured - answer-What are the two primary goals of a survey? Open-ended (participants generate a response) - more detail -allows for unforeseeable options - time-intensive Close-ended (participants chose from responses) - uniform responses allow for easy comparability, quick analysis - answer-What are the two broad categories of questions you often see on surveys? 1. Be clear 2. Keep questions short 3. avoid negatives 4. avoid double barrelled questions 5. Avoid leaning/biased questions 6. avoid asking questions that are difficult to answer 7. avoid false premises 8. avoid asking embarrassing questions unless necessary 9. Response options should -match questions- be mutually exclusive - be exhaustive - have an equal number of positive/negative responses - answer-9 survey question guidelines: over-reporting desirable behaviors, under reporting undesirable behaviors a. emphasis anonymity b. phrase questions in a way to "save face" - answer-What is social desirability bias? How do you overcome it? 1. Introduction of yourself and your purpose 2. Risk, benefits, and time required 3. Emphasis that you're not selling anything 4. Clear directions - answer-What 4 things should you include in the introduction to your survey? A skip pattern - answer-What is this an example of: "are you a graduating senior?" a. yes b. no If no, skip to question 3 NO - answer-In a survey should you use a wide variety of question styles? group them together - answer-In a survey, what should you do with similar question styles? Will questions you're asking influence questions that follow? - answer-What are order effects?