MKT 335 quizzes for Exam 2

What are the three rates one has to be mindful of that impacts the number of invitations you ultimately have to send to get your desired sample size

-Completion rate -Response rate -Incidence rate

Why do we prefer close-ended questions in general?

-Easier for participants to answer -Easier for researchers to code responses -Simpler wording

If content validity assesses whether the measure "covers" the full range of the research construct, what 2 dimensions must be captured for measuring satisfaction

-Expectations -Performance specifics

The motivations that underlie behaviors can be difficult to measure because

-Respondents may not be able to articulate why they behave the way they do -Respondents may be able to offer an explanation but it may not be the real reason for their behavior

In questionnaire design, a dummy table is most useful to help the researcher

-Think about the research questions without data -Pre-plan inferential data analysis

Optimal sequencing for the order and content of questions in a questionnaire is

1.) Broad 2.) Mid-Range 3.) Specific

Surveys that take less than 5 minutes to complete can expect approximately what percentage of respondents to not complete the survey?

20%

Which one of the following is an incorrect statement?

A sample cannot be too large

Key advantages of a questionnaire compared to observations include: (select all that apply)

Can get inside consumers' heads Versatility in types of data Is more structured

Which two of the following statements accurately describe sampling error?

Cannot be eliminated entirely We consider it "good error" because we can estimate it and control for it

What kind of validity will the professor push you to develop with your survey?

Content

As the number of respondents increases in your sample, what one thing below also typically increases? Select 1 below

Cost of the sample

What's wrong with the following survey statement? Do you like having Professor Bosarge and the flipped classroom? (NOTE: For the purpose of the question, ignore the fact there is no scale, just focus on the survey statement)

Double-Barreled Question

Your employer, located in 3 countries with 8 offices, has asked you to collect employee satisfaction data among their 2,200 employees. How would you best describe this population? Select 1 below

Finite

Which one of the following four characteristics is least likely to be used to describe a good sample?

Homogenous

What's the problem with a sample that's too large?

It is costly to collect data, so we are wasting resources

Why don't we always use the highest level of measurement? (select all that apply)

It's not appropriate for some variables (e.g. categories) It may make the question more difficult for respondents

Which of the following is not a level of measurement?

Likert

An open-ended question is best used when

Many variables/variants to assess Response comparisons are not important Deeper consumer context is needed

Which of the following is not a measure of central tendency?

Maximum

Validity is best and most simply defined as

Measuring what we intend to measure

What measures of central tendency can we use on Ordinal data? (select all that apply)

Median Mode

The level of measurement that best defines demographics data such as ethnicity and gender is

Nominal

What is wrong with this response scale? (select all that apply) How many siblings do you have? 1-2. 3-4. 2-3

Options are not mutually exclusive Options are not exhaustive Options are not in a logical order

Why is it better asking demographic questions at the end of your survey?

People will be more comfortable answering them after you have established rapport

All of the following are types of probability samples except:

Quota sample

What type of sample is it when you ask respondents to refer you to other possible respondents?

Snowball Sample

Which two of the following statements related to assessing reliability are correct?

Test/retest is a method used to test the same group of people twice with the same questions Cronbach's coefficient alpha is the measure of correlation among multiple items that measure the same construct

What is the sampling frame?

The set of all people that can be sampled

In questionnaire design, a dummy table is most useful to help the researcher (select all that apply):

Think about the research questions without data Pre-plan inferential data analysis Visualize question comparisons

A measure can be reliable but not valid

True

Why is sampling error sometimes called "good error"? (Select all that apply)

We can estimate how much of it there is We can reduce it by increasing sample size

Which of the following statements is correct regarding using composite items to measure constructs in a survey?

We should use composite items for constructs that are multi-dimensional

What of the following questions is most likely to need an open-ended response?

Why do you prefer your current shampoo brand?

Because they may be considered more personal questions, we typically place demographic questions at the _________ of a survey

end

A Stratified Sample is a non-probability sample

false

Reliability means actually measuring what we intended to measure.

false

Test/Re-Test is a form of Validity testing for Stability of data collection.

false

Your employer, located in 3 countries with 8 offices, has asked you to collect employee satisfaction data among their 2,200 employees. How would you best describe this population?

finite

To make our parameter estimates twice as precise, we would need a sample

four times as large

What is an example of Psychographics (Lifestyle) characteristics (select the one best answer)

interests opinions activities

The ACT and SAT exams are examples of which level of measurement?

interval

Our first priority in asking a question is to ask it in a way that

is easiest for participants to respond to

Convergent validity requires

multiple measures

Which one of the following four statements is incorrect regarding probability sampling procedure?

one can control for random sampling error even without probability sampling procedure

The level of measurement that best defines "ranked data" is

ordinal

If a researcher conducted a census of the entire population, they could calculate

parameters

What measure of dispersion is most appropriate for results to a ranking question?

percentile

All of the following are types of probability samples except

quota sample

Random Error is a threat to

reliability

The three necessary characteristics of a good sample are

representative accessible low cost

Why are researcher degrees of freedom a problem?

researchers can intentionally or accidentally introduce bias

Which of the following IS NOT one of the three broad types of questions typically asked in a survey?

satisfaction

Which is a method to assess reliability of a measurement?

test/retest alternate forms cronbach's coefficient alpha

A measure can be reliable but not valid.

true

Likert Scale questions are examples of Interval level of measurement.

true

One reason for collecting demographics and socio-economic characteristics is segmentation/targeting

true

Reliability is best and most simply defined as consistency of measurement over time

true

Unaided recall is one of the best measures of awareness

true

When is it appropriate to use an open-ended question on a survey?

when you can't list all possible responses