Market research and analysis

Question 1

What is the measurement process?

Answer 1

1. Identify the concept of interest (e.g. how ‘role ambiguity’ affects sales persons performance, note: abstract constructs existing in mind of the person); Use the concept to
2. Develop a construct; Which is used to create
3. A constitutive definition; Which enables a researcher to develop
4. An operational definition (what exactly do we mean by ‘role ambiguity’?); Which enables a researcher to create
5. A measurement scale; That requires the researcher to
6. Evaluate the reliability and validity of the scale; If the evaluation is satisfactory, the researcher
7. Utilizes the scale; Which leads to
8. Research Findings

Question 2

What are the levels of measurement?

Answer 2

The levels of measurement are:

1. Nominal: Scales that partition data into mutually exclusive and collectively exhaustive categories (What is your gender?)
2. Ordinal: Scales that maintain the labeling characteristics of nominal scales and have the ability to order data. (Put these 5 brands in order from your most favorite to least favorite.)
3. Interval: Scales that have the characteristics of ordinal scales, plus equal intervals between points to show relative amounts; they MAY include an arbitrary 0 point (like a thermometer does). (E.g. Rate on a scale of 1 to 5 how much you agree or disagree with the following statement.)
4. Ratio: Scales that have the characteristics of interval scales, plus a meaningful 0 point so that magnitudes can be compared arithmetically. (What is your income?)

The data you collect determines the allowable analysis.

Question 3

What is nominal data?

Answer 3

Nominal data: Scales that partition data into mutually exclusive and collectively exhaustive categories. (Can't be one if your the other)
Classification type data:
- Gender
- race / ethnicity
- occupation
Analysis approach:
- cross tabulations / percentages
- sums and frequency counts

Question 4

What is ordinal data?

Answer 4

Ordinal data: Scales that have the ability to ORDER data.
Ranking type data:
- best liked, worst liked
- 1st, 2nd, 3rd
- comparison rankings (rank these movies best to worst)
Analysis approach:
- cross tabs, sums and frequency counts
- percentages, mode (most occurring value)

Ranking is mentally easy and tends to be stable over time, but is limited in how it can be analysed.

Question 5

What is interval data?

Answer 5

Interval data: Scales that have the characteristics of ordinal scales, plus equal intervals between points.
Comparison type data:
- ‘1 to 7’ or ‘-3 to +3’ scales, which are commonly used: how strongly do you agree with the following statements?
- age, income etc. can be put into equal ranges
Analysis approach:
- standard deviation, skewness, kurtosis
- descriptive stats. (mean, median, mode)
- frequency counts
- correlation

Question 6

What is Ratio data?

Answer 6

Ratio data: Scales that have the characteristics of interval, plus a meaningful 0 point.
Flat numeric type data:
- What is your age?
- Income?
Analysis approach:
- standard deviation, variance, kurtosis
- desc. stats (mean median mode)
- frequency counts
- correlation, regression

Question 7

What is reliability vs validity?

Answer 7

Reliability = degree to which measures are free from random error and, therefore, provide consistent data.
Validity = Degree to which what the researcher was trying to measure was actually measured.

Question 8

What are the different ways to test reliability? (4)

Answer 8

1. Test and retest: ability of same instrument to produce consistent results when used a second time under conditions as similar as possible to original conditions.
2. Equivalent form: ability of 2 very similar forms of an instrument to produce closely correlated results.
3. Internal consistency: ability of instrument to produce similar results when used on different samples during same time period to measure a phenomenon/
4. Split Half Technique: assessing reliability of scale by dividing total set of measurement items in half and correlating results.

Question 9

What are the different ways to test validity? (4)

Answer 9

1. Face validity: degree of measurement seems to measure what is supposed to be measured.
2. Content: representativeness of content of measurement instrument
3. Criterion related: degree to which measurement instrument can predict a variable of interest
4. Construct: degree to which a measurement instrument represents and logically connects observed phenomenon to construct.

Question 10

What is a RATING SCALE?

Answer 10

Rating scale = respondent selects an answer from a limited number of ordered categories.
Odd scale: (1-5): 
- Can be appealing to respondents since there is an easy option to select
- If topic is highly sensitive, may be best to offer neutral point
Even scale (1-6)

Question 11

What is a LIKERT SCALE?

Answer 11

A Likert scale is a 5 point interval scale (these are easy to answer and very common)
E.g. The registration was simple. Strongly disagree - neutral - strongly agree

Question 12

What is a SEMANTIC DIFFERENTIAL SCALE?

Answer 12

Like Likert scales, 5 or 7 points are common. Difference is the scale is the question, not the response to a statement. E.g. 1 = modern, 7 = old-fashioned.

Question 13

What is a GRAPHIC RATING SCALE?

Answer 13

These scales are often used when interviewing children. This is an interval scale. Uses images. E.g. sad faces and smiley faces. If they circle the smiley face, we assign the number 1.
E.g. 2: measurement scales that include a graphic continuum, anchored by two extremes. (A = uncomfortable to comfortable, cross on the line where you fit) (B = same, but with markers) (C = graph of thermometer increasing as comfortability increases.

Question 14

What is a RANK ORDER SCALE?

Answer 14

List 6 answers against a couple of questions and ask that the best be ranked 1 and the worst be ranked 6.

Question 15

What is a PAIRED COMPARISON?

Answer 15

Putting 2 characteristics against each other in a pair and asking that the best be chosen in each pair. (Could plot frequency count of number of times ‘A’ was selected)

Question 16

What is a CONSTANT SUM SCALE?

Answer 16

Ranking a group of characteristics adding up to 100 points in order of importance. E.g. 7 points = 10, 20, 50, 5, 5, 5, 5 = 100. (People have problems getting numbers to add to 100, certainly with 7 options, but more insight is gained into relative value of features.

Question 17

What is STAPLE SCALE?

Answer 17

Place the statement in the center of a + and - scale (up to 5) and the respondent can show how much they either agree or disagree with the statement.

Question 18

What is a PURCHASE INTENT SCALE?

Answer 18

Purchase intent scales find out your intent in regard to a certain product release etc. You simply circle which you would probably do. E.g. definitely buy proposed product, probably buy, probs not, and defs not.

Question 19

How do you select the correct scale?

Answer 19

1. The nature of the construct being measured (is it an observable construct?)
2. Type of scale and number of scale categories. (ease of development and execution are certainly factors. Measurement scales exist, but use with caution)
3. Balanced vs Non-balanced.
   - Balanced = scales with equal numbers of positive and negative categories.
   - Non- balanced = scales weighted towards one end or the other of the scale
4. Forced vs non-forced. Having an odd vs even number of response choices.

Question 20

What is the QUESTIONNAIRE DESIGN PROCESS?

Answer 20

1. Determine survey objectives, resources, and constraints
2. determine the data collection method
3. determine the question response format
4. decide on the question wording
5. establish questionnaire flow and layout
6. evaluate the questionnaire
7. obtain approval of all relevant parties
8. pretest and revise
9. prepare final copy
10. implement the survey

Question 21

What are some tips for writing a good questionnaire?

Answer 21

1. avoid abbreviations, slang, or uncommon words that your audience might not understand
2. be specific. vague questions generate vague answers. on the other hand, don’t overdo it
3. make sure your questions are easy to answer
4. don’t assume to much
5. watch out for double questions and double negatives
6. check for unintentionally creating bias

Question 22

What is coding?

Answer 22

The process of grouping and assigning numeric codes to the various responses to a question. I take the first completed questionnaire and call it my ‘coding sheet’. All responses to close-ended questions are assigned a number and we endeavour to put open-ended responses into groups.

Question 23

What are the survey objectives, resources, and constraints?

Answer 23

Objectives - outline of the decision making information required
Resources - budget in terms of money, time, and personnel
Constraints - budget? time?

Question 24

Why is the data collection method important?

Answer 24

The data collection method will have a major impact on the questionnaire design and the projects time and money budget
E.g. in person (small intercepts), telephone, mail, internet.
Very different completing survey by yourself vs someone reading qu’s to you.

Question 25

What is the response format?

Answer 25

open ended: Questions to which the respondent replies in his or her own words. A means to gain insight into consumer vocabulary, but use these sparingly. Among other things (like coding), it is mentally effortful.
Closed ended: Questions requiring respondents to choose from a list of answers.

Question 26

How do you get the survey flow and layout?

Answer 26

Screeners: ask general questions first, establish respondent ‘buy in’. (Have you shopped at __ in the past month?)
Warm ups: gets the respondent thinking about the topic at hand, establishes parameters about respondents attitudes, behavior etc
Transitions: Questions that set the tone for the more difficult or specific qu’s to come.
Complicated: Use of rating scales for attributes, attitudes, beliefs, opinion etc (tackling controversial issues)
Classification: personal or demographic qu’s (more invasive)

Question 27

What question do you need to ask to evaluate a survey?

Answer 27

• Is the question necessary?
• Is the questionnaire too long?
• Will the questions provide the information needed to accomplish the research objectives?

Question 28

What are the pros and cons of internet surveys?

Answer 28

pros: The questionnaire’s appearance consistency is easier to achieve.
The questionnaire can be checked for typos easily.
Skip patterns can be efficiently established.
The survey can be distributed quickly for expert review & input.
Cons: Over reliance on electronic survey construction can lead to the researcher’s getting sloppy as he/she might think the software will do the work and correct any errors.
The researcher might feel less connected to the process.
Multiple versions of the survey might get circulated / distributed.

Question 29

How do I develop a sampling plan?

Answer 29

1. Define the population of interest (Determine the characteristics of those you are interested in studying)
2. choose a data-collection method (Determine how you collect the sample - such as mail, Internet)
3. identify a sampling frame (A list of population elements from which units to be sampled can be selected)
4. select a sampling method (Determine how you will get the sample, whether through probability or non-probability methods. Sometimes we crudely collect a probability sample)
5. determine sample size (What is the level of accuracy you want to achieve, the time and money you have to do the survey, and the data collection method?)
6. develop operational procedures for selecting sample elements (This is the plan of how to go about actually choosing and interviewing the respondents.)
7. execute the operational sampling plan (Field workers must be trained to execute the sampling plan properly.)

Question 30

How do you gather a probability sample?

Answer 30

Simple random: A sample selected by assigning a number to every element of the population and then using some method for randomly selecting elements to be in the sample such as random digit dialing. (sample size / pop size)
Systematic: A (close enough to a random) sample in which the entire population is numbered and elements are selected using a skip interval – every “nth” name is selected (pop size / sample size)
Stratified: A sample that is forced to be more representative through simple random sampling of mutually exclusive and exhaustive subsets either proportionally or disproportionally.
Cluster: A sample in which the sampling units are selected from a number of small geographic areas to reduce data collection costs

Question 31

How do you gather a non probability sample?

Answer 31

Convenience: A sample based on using people who are easily accessible (assessment)
Judgement: A sample in which the selection criteria are based on the researcher’s personal judgment about the representativeness of the population under study.
Quota: A sample in which quotas, based on demographic or classification factors selected by the researcher, are established for population subgroups.
Snowball: A sample in which additional respondents are selected based on referrals from initial respondents.

Question 32

What are the pros and cons of internet sampling?

Answer 32

Pros: Target respondents can complete the survey at their convenience.
Data collection is inexpensive.
Survey software can facilitate the data collection process.
The survey can be completed quickly.
Cons: Sample might not be representative of the population – for starters, they have to internet users.
You cannot always be sure who is completing the survey.
Maintaining respondent confidentially can be a challenge, but there are ways to address this.
Data security issues can be difficult to manage, but there are ways to address this.

Question 33

How do you determine your sample size?

Answer 33

Budget available. Rule of thumb = 100 per group for quan surveys, 20-30 for qual studies

Question 34

How many contacts do you need?

Answer 34

you need to contact more than n (your desired sample), since its unlikely that everyone you contact will agree to answer the survey. if you need 800 respondents and typical response rate = 30%, then you would actually need to contact 2667 ppl to get 800 completed surveys.
n/.3 (800/.3) = 2667

Question 35

What is the normal distribution?

Answer 35

The continuous distribution that is bell shaped and symmetrical about the mean. The mean, median, and mode are equal. Finally, about 68% of the observations are within one standard deviation plus/minus of the mean, ~95% are within two standard deviations, and 99+% are within three standard deviations of the mean, respectively.

Question 36

What is the standard deviation?

Answer 36

The square root of the variance, calculated by taking away each unit from the mean (calc. by adding all together dividng by how much)

Question 37

What is the central limit theorum?

Answer 37

The idea that a distribution of a large number of sample means will approximate a normal distribution - regardless of the distribution of the population from which they were drawn!