Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
129 Cards in this Set
- Front
- Back
The process of using rules and procedures to assign numbers to objects (or entities or constructs)to represent quantities of properties of those objects
|
measurement
|
|
Defining something by the rules or operations used to assign numbers to a property
|
operational definition
|
|
what we measure
|
concept
|
|
concept that is carefully defined for use in research
|
construct
|
|
This characteristic or "property" of a person can take different values such as low or high brand loyalty, we often refer to it as
|
variable
|
|
The description of a property in words
|
constitutive definition
|
|
when we say what operations are needed to measure a construct (such as having a person make a mark on a paper questionnaire)
|
operational definition
|
|
respondents are instructed to respond in their own words
|
open-ended question
|
|
provides options on the questionnaire that can be answered quickly and easily
|
closed-ended question
|
|
special questions used to measure properties (constructs) on an underlying continuum of intensity
|
scaled response question
|
|
Four Considerations in Choosing a Question-Response Format:
|
1. Ability of the Respondent
2. Nature of the Property Being Measured 3. Previous Research Studies 4. Scale Level Desired |
|
3 Strengths of Open-Ended Questions:
|
1. Respondents can use own words
2. Researcher doesn't have to think of answers ahead of time 3. Can yield qualitative information |
|
2 Weaknesses of Open-Ended Questions:
|
1. Hard to code, interpret, and analyze
2. Respondents may do poor job if they have to write their own answers |
|
3 Strengths of Closed-Ended Questions:
|
1. Easy to ask
2. Easy to answer 3. Easy to code |
|
2 Weaknesses of Closed-Ended Questions:
|
1. Over-simplify response options
2. Don't even allow a "don't know" |
|
defined as those that use only labels; that is, the possess only the characteristic of description
|
nominal scales
|
|
permit the researcher to rank-order the respondents or their responses
|
ordinal scales
|
|
those in which the distance between each descriptor is known
|
interval scales
|
|
ones in which a true zero origin exists - such as an actual number of purchases in a certain time period, dollars spent, miles traveled, number of children in the household, or years of college education
|
ratio scales
|
|
Always measure a construct at the _______ level of information possible.
|
highest
|
|
a form of an itemized rating scale where the numbers are given labels of agreement level
|
Likert scale
|
|
are often used for store image measures and sometimes attitudes
|
semantic differential scales
|
|
Why are the positive and negative ends of the scale "flipped" back and forth in questionnaires?
|
To avoid the halo effect
|
|
a general feeling about a store or brand that can bias a respondent's impressions of its specific properties
|
halo effect
|
|
Any question or scale needs ___________ and _________.
|
reliability; validity
|
|
Reliability and validity
Which can you have without the other? |
Can be reliable but not valid
|
|
Reliability and validity
Which can you NOT have without the other? |
Cannot be valid if it is not reliable
|
|
a multiple part measuring tool
|
questionnaire
|
|
Eight Steps of Questionnaire Development:
|
1. Specify the survey objectives
2. Determine type of questionnaire 3. Develop the individual question content 4. Determine the method of response to each question 5. Develop the individual question wording 6. Question sequencing in the questionnaire 7. Work on the physical appearance of the questionnaire 8. After client approval, pretest the questionnaire |
|
What information is sought when specifying the survey objectives?
|
List the constructs (name the variables) and the more specific properties measured for complex constructs
|
|
It is always preferable to use ____________ questions if possible.
|
closed-ended
|
|
Ten Words to Avoid in Question Development:
|
1. All
2. Always 3. Any 4. Anybody 5. Best 6. Ever 7. Every 8. Most 9. Never 10. Worst |
|
simplest and uses a wide-to-narrow or general-to-specific flow of questions that places general inquiries at the beginning of a topic on the questionnaire and those requiring more specific and detailed responses later on
|
funnel approach
|
|
employed when the researcher realizes that respondents will need to apply a different amount of mental effort to various blocks of questions
|
work approach
|
|
organizes questions into sets based on the common objective of the questions in the set
|
sections approach
|
|
8 Common Mistakes in Developing Entire Questionnaires:
|
1. Asking unnecessary questions
2. Giving respondents no incentives 3. Asking "silly" questions which have obvious answers 4. Asking loaded or leading questions 5. Using ambiguous wording 6. Using too few response categories 7. Breaking promises to respondents 8. Using too long or complicated forms |
|
any error in a survey due to the fact that a sample is used and is caused by two factors
|
sampling error
|
|
it’s caused by chance variation in the particular sample members you select
|
random sampling error
|
|
means that population lists are not representative
|
sampling frame error
|
|
95% of all research error
|
non-sampling error
|
|
caused by the interviewing or observation processes used
|
observation error
|
|
caused by the failure to get the interview successfully completed
|
non-observation error
|
|
Three Types of Observation Error:
|
1. Fieldwork
2. Respondent 3. Caused by the situation |
|
Three Types of Non-Observation Error:
|
1. Non-contact
2. Refusal 3. Wrong target |
|
means removing all of the mistakes from a data file so it can be used in analysis
|
data cleaning
|
|
means the assignment of numbers to types of meanings in a response so that the data can be manipulated with a computer
|
data coding
|
|
the list of variables with the associated code numbers
|
code book
|
|
application of statistical analysis to data to find, summarize, and communicate information about the data
|
data reduction
|
|
Recall that the manager needs information for _______, not just "raw" data
|
decisions
|
|
Five Types of Statistical Analysis:
|
1. Descriptive
2. Inferential 3. Difference Tests 4. Associative 5. Predictive |
|
________ _________ are useful for descriptive analysis.
|
frequency distributions
|
|
tabulation of the number of times that each different value appears in a particular set of values
|
frequency distribution
|
|
arithmetic average of a set of numbers
|
mean
|
|
the value whose occurrence lies in the middle of a set of ordered values
|
median
|
|
identifies the maximum and minimum values in a set of numbers
|
range
|
|
indicates the degree of variation in a way that can be translated into a bell-shaped curve distribution
|
standard deviation
|
|
Why is the "standard deviation" instead of the range used to show the variation in a set of data?
|
The Range shows the lowest and highest scores but doesn't show how the observations may or may not be clumped around the mean
|
|
If we say that “Kevin is highly intelligent because his IQ is 125,” which type of definition of the property "intelligence" are we using?
dictionary constitutive operational construct |
operational
|
|
The term used for a property that is scientifically defined and measured by a question on a questionnaire is...
scale definition concept construct |
construct
|
|
When a construct that we measure has different values for different respondents, it varies among people. That is why we refer to it as a…
concept variable concept definition |
variable
|
|
What makes an open-ended question difficult to use with mailed or computer delivered questionnaires?
They are hard to code. (put into numbers) They are hard to analyze. Some people can’t or won’t answer. All of these are true |
All of these are true
|
|
There are three basketball players with numbers 2, 4 and 10. Which is true?
#4 is more skillful than #2. #10 is five times better than #2 The difference in skill between #2 vs #4 is the same. You can find their names in the program |
You can find their names in the program
|
|
“New Sudso detergent is gentler to your hands as the leading brand Raspo!” What level (kind) of scale is assumed?
Nominal Ordinal Interval Ratio |
Ordinal
|
|
“New Sudso detergent is TWICE as gentle to your hands as the leading brand Raspo!” What level (kind) of scale is assumed?
Nominal Ordinal Interval Ratio |
Ratio
|
|
If you believe your measurement scale is measuring a person's health orientation but really is measuring a person's tendency to give desirable answers, you have what type of problem?
reliability validity word choice formatting |
validity
|
|
What type of data is represented by the dollar price of bolts?
Nominal Ordinal Interval Ratio |
Ratio
|
|
Last year my firm charged $35 for a large machine bolt and now charges $70. How much has the price gone up?
Fifty percent Doubled...twice the price Infinite I don’t have a calculator |
Doubled...twice the price
|
|
What type of data is represented by the temperature in degrees fahrenheit?
Nominal Ordinal Interval Ratio |
Interval
|
|
If it was 35o yesterday but today it is 70o. How much warmer is it today than yesterday?
35o Twice as warm All of the above None of the above |
35o
|
|
Is the difference in my preference for corn and peas the same as the difference in my preference for broccoli and parsnips?
Yes No Maybe-can’t tell |
Maybe-can't tell
|
|
Considering the 5 vegetables that I ranked, do I like broccoli?
Yes No Can’t tell |
Can't tell
|
|
If I asked all of the students in the class to rank the 5 vegetables, could I calculate a meaningful average ranking of their preferences?
Yes No Maybe Can’t tell |
No
|
|
Considering the 5 vegetables that I rated, do I like broccoli?
Yes No Can’t tell |
Yes
|
|
If I asked all of the students in the class to rate the 5 vegetables, could I calculate a meaningful average rating of their preferences?
Yes No Maybe Can’t tell |
Yes
|
|
Is the difference in my preference for corn and peas the same as the difference in my preference for rutabaga and parsnips?
Yes No Maybe-can’t tell |
Yes
|
|
Do I like parsnips twice as much as I like Rutabaga?
Yes, since 2 is twice as big as 1 Definitely No Don’t know--This is an interval scale so you can’t make a ratio with the numbers This is a ratio scale, so you can say twice as much |
Don’t know--This is an interval scale so you can’t make a ratio with the numbers
|
|
What level of data did the previous question provide to the researcher? (be careful in considering this one!)
Nominal Ordinal Interval Ratio |
Nominal
|
|
What type of question is the following?
Why were you chosen to participate in the selection of an accounting firm? _______________________________ Closed ended – dichotomous Closed ended – multiple category Open ended – unprobed Open ended – probed Scaled response |
Open ended – unprobed
|
|
What level of data did the previous question provide to the researcher? (be careful in considering this one!)
Nominal Ordinal Interval Ratio |
Nominal
|
|
If a question yields the same answer from a person on two different occasions (assuming the person and the situation have not changed) we say the question is…
valid reliable consistent dependable |
reliable
|
|
Considering Step #1: Is the Question Necessary?Assume you are doing a study of consumer satisfaction with Coleman camping lanterns. While you are developing the questionnaire, the client says: “As long as your people are on the phone, why don’t they ask what foods campers like to cook at camp?” Would you say...
Yes, no problem Yes, but it will cost extra. No, it will require more interviewers No, it could mess up our satisfaction study. |
No, it could mess up our satisfaction study.
|
|
Considering Step #3 -- Are several questions needed to get data for the information?Assume you are doing a study of consumers’ image of the Ford Motor Company. Knowing what you have learned in marketing about brand image, how many items (questions) would you need?
One question (good vs. bad image) A series of items on different attributes of image Two items (a) good vs. bad and (b) intensity of feelings |
A series of items on different attributes of image
|
|
If respondents answer a question incorrectly because they are worried about what the interviewer might think of them, we would say the question is…
Inaccurate Not reliable A waste of time Invalid |
Invalid
|
|
How would you contact a respondent to ask question #15a (in the previous slide) most effectively?
Telephone Face-to-face Computer |
Computer
|
|
How would you contact a respondent to ask question #19 most effectively?
telephone computer |
telephone
|
|
“What year did you graduate from high school?” could be used in place of the first part of question #21. As such, what type of question is it?
Third person technique Disguised question “Normalized behavior question Hidden question Randomized response question |
Disguised question
|
|
What’s wrong with this question?“How often do you drink regular caffeinated coffee? A) Extremely often B) Very often C) Not too Often D) Rarely
Not one topic Not short enough Poor sentence structure Ambiguous wording |
Ambiguous wording
|
|
Respondents are greatly influenced by the alternative answer options in a question. When a question or its options are not fully described or not complete, the respondent will assume alternatives. Thus there are “implicit alternatives.” Which question has an implicit or unstated alternative? (In the 1960’s women gave it the most answers indicating a desire for a job.)
Would you like to have a paid job if it were possible? ___yes ___no 2. Would you like to have a paid job if it were possible or do you prefer to do just your housework? ____Prefer job as well as my housework ____Prefer housework only |
Would you like to have a paid job if it were possible?
___yes ___no |
|
What is wrong with this question?When you buy fast food, what percent of the time do you order French fries?
Asks for a broad generalization Sensitive question Leading question Multiple topic question |
Asks for a broad generalization
|
|
What is wrong with this question? “Should Taco Bell, a firm that exploits vulnerable migrant farm workers, be banned from the Kirkhof Center?” The question is…
Leading Loaded Too general One sided |
Loaded
|
|
What’s wrong with this question? “What kind of car is do you believe environmentally concerned consumers should own?"
Use of biasing word Asking for generality Ambiguous word or phrase Poor sentence structure |
Poor sentence structure
|
|
What’s wrong with this question? Don’t you agree that, as Al Gore says, electric cars are the answer to the carbon emissions problems faced by the United States?
Loaded question Asking for generality Leading question Poor sentence structure |
Leading question
|
|
What is wrong with this question? Should the State of Michigan allow people to care for other people’s children for up to two hours per day without getting a day care license? The question is…
Leading Assumes criteria that are not clear Loaded One sided |
Assumes criteria that are not clear
|
|
What’s wrong with this question? “Was the “cash for clunkers” program an egregious government budget expenditure?”_____Yes _____No _____Don’t remember
Use of biasing words Asking for generalities A word is not core vocabulary Double barreled question |
A word is not core vocabulary
|
|
Did you consider all brands of chips when you selected snacks for your most recent party?_____Yes _____No _____Don’t remember
Use of biasing words Asking for generalities A word is not core vocabulary Dead give away word. |
Dead give away word
|
|
What’s wrong with this question? “How do you feel about the budget busting Obamacare law and all of its requirements?
Use of biasing words Asking for generalities Loaded question Double barreled question. All of the above |
All of the above
|
|
Identify the error in the question: “Should we put a lid on new big-government programs and instead give hard working Americans a tax cut?"
Leading Ambiguous words Dead give-away words Asks for specifics - gets guess at generality Loaded Implicit assumptions |
Loaded
|
|
When a numerical response is needed, it is usually best to ask respondents to state a number. So why would the following question be appropriately designed?What is your household’s annual income after taxes?____Under $25,000 ____$26,000-$50,000 ____$51,000-$75,000 ____Over $75,000
Leading Ambiguous words Dead give-away words Asks specific - gets generality Leading Sensitive Issue |
Sensitive Issue
|
|
Identify the error in this question: "Should the state provide private school tuition vouchers and similar programs?"
Leading Ambiguous words Dead give-away words Asks specific - gets generality Loaded Implicit assumptions |
Implicit Assumptions
|
|
Identify the error in this question: "Should the state give parents free choice in getting their children a quality education by providing private school tuition vouchers if it means taking money away from public schools?"
Loaded Ambiguous words Dead give-away words Asks specific - gets generality Leading Implicit assumptions |
Loaded
|
|
Identify the error in this question:"How often do you record TV programs for later viewing? ___Rarely ___Occasionally ___Sometimes ___Often
Double barreled Ambiguous wording Dead give-away words Asks specific - gets a generality Leading Implicit assumptions |
Ambiguous Wording
|
|
Identify the error in this question to a purchasing agent: "How many salespeople did you see yesterday?
Double barreled Ambiguous word Dead give-away word Asks specific - gets a generality Leading Implicit assumptions |
Ambiguous Wording
|
|
Identify the error in this question to a purchasing agent: "How many salespeople did you personally talk with last year?
Double barreled Ambiguous word Dead give-away words Asks specific - gets a generality Leading Implicit assumptions |
Asks specific - gets a generality
|
|
What is the error in this questionto professors?"When I order text books from the university store, they always get enough for my class" ____agree ____disagree
Double barreled Ambiguous word Dead give-away word Asks for generalization Leading Implicit assumption |
Dead give-away word
|
|
In the 1936 election, the Literary Digest wrongly picked Republican Alf Landon to beat Franklin Roosevelt. Why? Because they had randomly sampled and called names from telephone books around the country. Was their error due to…
Random sampling error Sampling frame error Questionnaire design error Non-sampling error |
sampling frame error
|
|
Assume that a population is 61% female. From a representative list of the population members we randomly draw a sample of 100 people and get only 52% females. What explains this?
Random Sampling error Sampling frame error Questionnaire design error Non-sampling error |
random sampling error
|
|
The leading cause of interviewer cheating is the…
Morals of the interviewers Lack of training Compensation system Lack of supervision of interviewers in the field |
compensation system
|
|
What appears to be the marketing research industry's greatest data collection problem?
Scarcity of trained interviewers Inability to locate respondents selected In a sample Difficulty in switching to on-line technology Growing respondent refusal rates |
Growing respondent refusal rates
|
|
How should you define the survey response rate? (you must tell your client)
(#complete) / (# in sample) (# complete) + (# refused)/(number in sample) (# contacted) / (# eligible in sample) (# complete)/ (# eligible in sample) |
(# complete)/ (# eligible in sample)
|
|
Calculate the response rate: Assume you called 500 people and you were only able to actually contact 400. Out of those 400 you get 200 completed interviews while 100 were not eligible for the study and 100 refused to be interviewed. Approximately what was your response rate as defined by CASRO?
40% 50% 60% 66% |
60%
|
|
In the fall 2005, the MKT352 professor randomly selected 40 classes from the GVSU schedule of classes. Of the 40 classes, only 7 were in business subjects downtown at the DeVos Center. Each class member was assigned one of those 40 classes and asked to interview four students as they left the class. A total of 52% of the interviews were done with business students. What occurred?
Random sampling error Sampling frame error Failure of field editing Lack of field controls |
Lack of field controls
|
|
What level of data is variable #1?1. How many credit hours are you enrolled in this semester? #_________ (Discontinue with thanks if fewer than 12 hours)
Nominal Ordinal Interval Ratio |
Ratio
|
|
What level of data is variable #2? 2. Do you know what the "wellness" concept is? _____ Yes _____ No (skip to #4)
Nominal Ordinal Interval Ratio |
Nominal
|
|
What level of data is variable #3? 3. In your own words, what definition comes to mind when you hear the term "wellness"? _______________________________________________________________ _______________________________________________________________
Nominal Ordinal Interval Ratio |
Nominal
|
|
How many different individual variables should be coded in this question?"Which food in the following list is your favorite one. Indicate it with a check mark?" ____ Apples____ Artichokes____ Bananas____ Baloney____ Zinfandel grapes____ Zuchini
one two four six |
One
|
|
What mistake is made in this question? How many retail stores belong to your chain of stores?____ 1 to 5____ 6 to 10____ Over 10
There should be more categories There should be narrower categories Categories should not be used There should be a “zero” category |
Categories should not be used
|
|
For how many variables must the person doing the coding use a non-zero number for the following answer given by one respondent?“Down the block there’s some people from Iran or Iraq or some place like that. I think they are Muslims ‘cause of what they wear.”
One variable Two variables Three variables Four variables |
Two variables
|
|
How many variables will have a zero entered in the computer record for the question in the previous slide? “Down the block there’s some people from Iran or Iraq or some place like that. I think they are Muslims ‘cause of what they wear.”
One variable Two variables Three variables Four variables |
One variable
|
|
How many variables are there in the question asked about sources of information?
One Six Seven Can’t tell |
Seven
|
|
Now how many variables are there in the question asked about sources of information?
One Six Seven Can’t tell |
One
|
|
How many variables are there in the question asked about satisfaction?
One Five Six Can’t tell |
One
|
|
When your questionnaire data has been collected, coded and entered in "raw" form in a file in columns and rows, we have a…
Data summary Data matrix Data analysis Data code book |
Data matrix
|
|
For what is the following formula?We say that this shows the "central tendency" of our data
Mean Standard Deviation Correlation Mode |
Mean
|
|
What level or type of data is V2, "heard of in newspaper"?
Nominal Ordinal Interval Ratio |
Nominal
|
|
Can we calculate a meaningful average score for V2?
No Yes Maybe It depends on how the interviewer asked the question |
No
|
|
What level or type of data is V9, "how satisfied are you with your membership"…?
Nominal Ordinal Interval Ratio |
Interval
|
|
Basing your answer on the previous two slides, for which is calculation of the mean nonsense?
V2 V9 Both V2 and V9 Neither V2 or V9 |
V2
|
|
For what is the following formula?
Mean Standard Error Correlation Standard Deviation |
Standard Deviation
|
|
The range and standard deviation are both considered…
Inferential statistics Associative statistics Measures of central tendency Measures of variation |
Measures of variation
|