exam 3 marketing research methods

0.0(0)
Studied by 0 people
call kaiCall Kai
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
GameKnowt Play
Card Sorting

1/52

encourage image

There's no tags or description

Looks like no tags are added yet.

53 Terms

1
New cards

The process of describing some characteristic or attribute of a phenomenon by assigning numbers in a consistent, organized way.

what is measurement

2
New cards
  1. Define what you are measuring: Decide exactly what you want to study or measure.

  2. Identify scales: Choose the correct way to measure what you need.

  • Scale of measurement: A system used to assign numbers or labels to what you are measuring.

  • Scale: A device or tool that includes a range of values used to measure something.


what are the steps of measurement

3
New cards

Choose the correct way to measure what you need.

what is a scale of measurement

4
New cards

scale range: Designated degrees of intensity assigned to responses to a question.

scale format: A set of descriptors used to represent the possible responses to a question (ex. centimeters or inches).

what are the different scales of measurement

5
New cards

unit: The unit used to measure something (ex. inches, centimeters, pounds).


point: A specific value or position on a scale.(the number you get)

what is the difference between a scale point and scale unit

6
New cards

nominal scale (assignment): Unique descriptors identify each object in a set. (what category)

ordinal scale (order): Shows the relative ranking or order of each scale point.(what order)

interval scale (distance): Shows the exact difference between scale point descriptors. (how far apart?)

ratio scale (origin): Has a unique starting point (true zero) in the set of scale point descriptors. (true zero)

what are the levels of scale measurement

7
New cards

-an individual is in either one category or another

-qualitative type of scale: Labels cannot be ordered because they have no meaning other than the number or word assigned to them.


names or categories only, there is no ranking or order

what happens during nominal scale measurement

8
New cards

-median or mode can be used


-ordinal= order or ranking

  • Shows which option is higher, lower, better, or preferred.

  • Shows the order but NOT the exact difference between choices.(you know you like mocha more than americano but you don’t know how much more)


what happens during ordinal scale measurement

9
New cards
  • know the exact distance/difference between scale points

  • mean and standard deviation can be used


internal=distance between numbers

  • Shows the exact difference between scale points.

  • The spaces between numbers are equal.

  • There is no true zero.


what happens in internal scale of measurement

10
New cards

-the highest form of measurement

ratio= everything an interval scale(exact distance) + a true zero (you did not do this at all)

what happens in ratio scale of measurement

11
New cards

Level

Main rule

What it means

Common statistics

Nominal

Assignment

=gender or yes/no question

Objects are either the same or different (categories).

Frequency, Mode

Ordinal

Order

=ranking coffee drinks from fav to least fav

Objects can be ranked from greater to smaller.

Frequency, Mode, Median, Range

Interval

Distance

=rating something 1-7

The spaces between numbers are equal.

Mean, Median, Variance, Standard Deviation

Ratio

Origin

number of times you ate at chipotle

Has a meaningful zero; can compare actual amounts.

Geometric Mean, Median, Variance, Standard Deviation


what are the levels of measurement and their common statistics

12
New cards

Respondents provide information by answering questions asked by a researcher.

Includes:

  • Rank: Put items in order.

  • Rate: Give something a score or rating.

  • Sort: Organize items into groups.

  • Choose: Select one answer from multiple choices.


what are self report scales

13
New cards

Asks respondents how much they agree or disagree with a series of statements.

Example:

  • Dunkin makes the best coffee.

  • I always get coffee from Dunkin.


Respondents answer on a scale from 1–7:
1 = Strongly Disagree
4 = Neutral
7 = Strongly Agree


pros: Can calculate scores based on the scale points chosen.

  • Good for capturing the intensity of an attitude (how strongly someone feels).


cons: Interpreting the results can be tricky.

  • Cannot infer the meaning behind someone's answer just from the number.


what is a likert scale and what are the pros and cons

14
New cards
<p>A <strong>rating scale</strong> that captures a person's attitude toward something.</p><ul><li><p>Measures a person's thoughts or feelings (cognitive or affective attitudes).</p></li><li><p>Only the endpoints are clearly defined.</p></li><li><p>Usually contains an odd number of scale points (such as 1–5 or 1–7).</p></li></ul><p></p><p>cons: Can be difficult to find words that are exact opposites.</p><ul><li><p>The neutral middle point can be hard to interpret.</p></li></ul><p></p>

A rating scale that captures a person's attitude toward something.

  • Measures a person's thoughts or feelings (cognitive or affective attitudes).

  • Only the endpoints are clearly defined.

  • Usually contains an odd number of scale points (such as 1–5 or 1–7).


cons: Can be difficult to find words that are exact opposites.

  • The neutral middle point can be hard to interpret.


what is a semantic differential scale and what are the pros and cons

15
New cards
<p>Measures how likely someone is to perform a certain behavior in the future.</p><p></p><p>pros : Good predictors of future behavior.</p><ul><li><p>Easy to create and use.</p></li></ul><p></p><p>cons:Hard to capture <strong>affective</strong> and cognitive components of attitudes.</p><ul><li><p>Labels may be unclear or mean different things to different people.</p><ul><li><p>Example: “Probably Would” could mean a 90–100% chance to one person but something different to another.</p></li></ul></li></ul><p></p>

Measures how likely someone is to perform a certain behavior in the future.


pros : Good predictors of future behavior.

  • Easy to create and use.


cons:Hard to capture affective and cognitive components of attitudes.

  • Labels may be unclear or mean different things to different people.

    • Example: “Probably Would” could mean a 90–100% chance to one person but something different to another.


what are behavior intention scales and what are the pros and cons

16
New cards

respondents make direct comparisons between two or more items and decide what they prefer

Example: “Which coffee do you prefer: Dunkin or Starbucks?”


pros: Provides fine-grained differences (shows small differences in preferences).

  • Easy to administer.


cons: May cause respondent fatigue if too many are used (people get tired of answering too many comparison questions).


what are comparative scales and what are the pros and cons

17
New cards

respondents rate something on its own without comparing different options. Example: “How much do you like this coffee?”

  1. Not at all

  2. Slightly

  3. Moderately

  4. Very much

  5. Extremely


pros: Measures someone's individual opinion or feeling about one item.

cons:More abstract and may be more difficult for respondents to answer.


what are noncomparative rating scales and what are the pros and cons

18
New cards
term image

what is a graphic rating scale

19
New cards

performance: expresses evaluative evaluations (judgments about how well something performs).

Example: Rate Cheesecake Factory's:

  • Food Quality: F, D, C, B, A, A+

  • Service Quality: F, D, C, B, A, A+


constant: respondents allocate points among different options.

Example: Allocate (set aside) 100 points among different brands based on how much you prefer each brand. All points must add up to 100.

what is the difference between performance rating scales and constant sum scales

20
New cards

allow people to compare options and express preferences.

Example: Rank TV shows from your favorite to least favorite:

  1. Top choice greys anatomy

  2. Second choice mad men

  3. Third choice game of thrones

  4. Fourth choice breaking bad


pros: Easy to administer.

cons:The number of choices should be small.

  • Too many options can make it difficult for respondents to rank everything because it requires too many comparisons.


what are rank order scales

21
New cards
<p>balanced scale: Equal number of positive and negative response options.</p><ul><li><p>Respondents can choose a neutral option.</p></li></ul><p></p><p>unbalanced scale: Unequal number of positive and negative response options.</p><ul><li><p>One side has more options than the other. </p></li></ul><p></p><ul><li><p>Want some variation, but not too much.</p></li><li><p>Want the scale to be easy to use and understand.</p></li></ul><p></p>

balanced scale: Equal number of positive and negative response options.

  • Respondents can choose a neutral option.


unbalanced scale: Unequal number of positive and negative response options.

  • One side has more options than the other.


  • Want some variation, but not too much.

  • Want the scale to be easy to use and understand.


what are the different design tips (2)

22
New cards

relevant questions: Ask questions that help solve the research problem.

  • Make sure each question connects to the construct or variable being measured.

  • Remove questions that do not help with the research objective, screening, or planned analysis.


accurate questions: Ask questions that collect valid data.

  • Answers should represent reality (what people actually think, feel, or do).

  • Pay close attention to how questions are worded and the order they appear in.


how do you decide what to ask in a questionnaire design

23
New cards

open ended questions: Respondents answer in their own words.

  • No answer choices are provided.

  • Allows people to explain their thoughts and opinions.

  • People write their own answers


fixed alternative questions: Respondents select one or more answers from the choices provided.

  • Answers are already listed for them.

  • Choose from the answers given


how do you decide how to word questions

24
New cards

pros: Respondents can express many thoughts and opinions.

  • Researcher does not limit the number or types of responses.

  • Provides detailed, information-rich answers.


cons: Takes more time to analyze.

  • Responses can be very different, making it harder to find patterns or common answers.


what are the pros and cons of unstructured open ended questions

25
New cards
<p>pros: Usually faster for respondents to answer.</p><ul><li><p>Data is easier and quicker to enter.</p></li><li><p>Makes analyzing results faster.</p></li></ul><p></p><p>cons: Potential bias because of:</p><ul><li><p>Order of the questions or answer choices.</p></li><li><p>The responses provided may influence what people choose.</p></li></ul><ul><li><p>Can be difficult if respondents have to think about too many options.</p></li></ul><p></p>

pros: Usually faster for respondents to answer.

  • Data is easier and quicker to enter.

  • Makes analyzing results faster.


cons: Potential bias because of:

  • Order of the questions or answer choices.

  • The responses provided may influence what people choose.

  • Can be difficult if respondents have to think about too many options.


what are the pros and cons of fixed alternative questions

26
New cards

avoid Leading Questions – questions that suggest or imply a certain answer, which can influence how respondents respond.

  • It may pressure respondents to answer "Yes."

  • This creates bias in the survey results.


keep it simple: Use words and grammar that respondents can easily understand.

  • Avoid complicated wording because it can cause confusion and incorrect answers.

  • Ask questions in clear, simple language.


avoid double barreled questions: questions that ask about two or more things at the same time.

  • Often contain words like "and" or "or."

  • Can confuse respondents because they may feel differently about each part.

  • One question should ask about one thing at a time


be specific with units of measure:

  • Clearly explain the time period or units you are asking about.

  • Avoid vague wording that could mean different things to different people.

  • Give specific answer choices so respondents know exactly how to respond

  • Be specific about when or how much so everyone understands the question the same way..


how should you phrase the questions you want to ask

27
New cards

apply the funnel technique: organize questions from general and easy to more specific and difficult.


  1. Screening Questions – make sure respondents qualify for the survey.

  2. Warm-Up Questions – general questions to get respondents comfortable.

  3. Focus Questions – more specific and detailed questions.

  4. Difficult Questions – keep respondents thinking consistently and answering carefully.

  5. Demographic Questions – ask about characteristics such as age, gender, and income.

  6. Thank You Statement – thank respondents for completing the survey.


how do you determine the order of questions

28
New cards

happens when the order of questions influences how someone answers.

what is order bias

29
New cards

physical appearance: Can Help or Hurt Your Survey!

  • Keep the survey as short as possible.

  • Include a Welcome Message explaining the purpose of the survey.

  • Use a simple background and clear fonts.

  • Make the survey easy to read and complete.


survey technology: Include a progress bar so respondents know how much of the survey is left.

  • Request a response from respondents.

  • Use randomization:

    • Randomize the order of questions.

    • Randomize the order of answer choices.


how do you check the layout of your questionnaire

30
New cards

trying out your survey before sending it to everyone.

  • Test the questionnaire before distributing it.

  • Have people similar to your intended respondents take the survey.

  • Ask them how they understood the questions and why they chose their answers.

  • Helps find mistakes and confusing questions before the actual survey is sent out

Look for:

  • Poor or unclear wording.

  • Missing answer choices or scale points.

  • Malfunctioning scales or survey features.


what is a pre test questionnaire

31
New cards

sample: Information collected from a subset (smaller group) of a population.

  • Measured using a sample.

  • Sampling allows businesses to make informed decisions without spending too much time or money.


population: Information collected from every member of a group.

  • Measured using a census.


what are sample basics

32
New cards

1. Define your target population

2. Select a sampling frame

3. Select a sampling procedure

4. Determine your sample size

5. Select individuals from your sampling

frame

6. Collect data

what are the sampling steps

33
New cards

target population:The specific group of people you are interested in studying.

  • The target population determines who you will collect your data from.

  • all Rowan University students.


elements: The individual people who provide the data.

  • Also called sampling units.

  • each individual Rowan student who is eligible to participate in the survey.


how do you define your target population

34
New cards

sampling frame: A list of potential people you can select from for your study (elements/sampling units).

  • It helps researchers identify who could be included in the sample.

  • Getting an accurate and complete sampling frame can be difficult and expensive.


examples) Phone book

  • Email directory

  • CRM database (customer information database)

  • Mailing list

  • Membership list

  • Online panels (groups of people who agree to participate in surveys)


how do you select a sampling frame

35
New cards

non probability sample: A sample where the probability of selecting each person is UNKNOWN.

  • Subjective: Researcher uses their judgment to choose participants.

  • Asking students who happen to be sitting in the student center to take a survey.


probability sample: A sample where each person has a KNOWN probability of being selected.

  • Objective: Participants are selected using a random process.

  • Randomly selecting 200 students from a complete list of Rowan students.


how do you select sampling procedure (2)

36
New cards

convivence sample: Selecting the people who are easiest or most convenient to reach.

  • Often includes people you already know.

  • Examples:

    • Asking friends or classmates.

    • Surveying people at a mall (mall intercepts).

    • Asking members of an organization.

problem: People in the same location or social group may have similar characteristics.


judgement sample: The researcher personally chooses respondents they believe meet the requirements of the study.

  • The researcher uses their own judgment to decide who should participate.

  • - Selecting certain cities to test a new product before launching it nationally (test-market cities).

problem: Researchers may think they are being unbiased, but their personal opinions or assumptions can influence who they select.


quota sample: A sample that tries to include people with certain characteristics in approximately the same proportions as the overall population.

  • The researcher sets a specific number (quota) of people needed from each group.

Example:

  • Population = 1,000 mall shoppers:

    • 620 females (62%)

    • 380 males (38%)

  • Desired sample = 100 people:

    • 62 females

    • 38 males


snowball sample: The researcher starts with a few people who meet the study's requirements and asks them to recommend other people who also qualify.

  • The sample grows as participants refer more people.

  • Example:

    • Researcher interviews 5 college students who work part-time.

    • Those students recommend other students who also work part-time.

problem: - May not be representative of the overall population because participants often recommend people similar to themselves.



what are the different types of non probability samples and their problems

37
New cards

Personal opinions and decisions influence who gets selected.

Bias is likely: The sample may favor certain types of people

There is no way to ensure the sample accurately represents the entire target population.

Results may be biased or lead to incorrect conclusions about the population.

what are the problems with non probability samples

38
New cards

simple random sample (SRS):Every sampling unit has an equal chance of being selected, like drawing names out of a hat

  • Participants are chosen randomly.

  • Usually uses a random number generator or random number table.


systematic random sample: Sampling units are put in an ordered list, and people are selected using a set interval.

  • First, choose a random starting point.

  • randomly pick a starting point, then select every certain number of people.

Example:

-Start with list of students

-Random starting point =Age 17 and k+=2

-Count every second person from age 17


stratified random sampling: he population is divided into similar groups called strata.

  • Random people selected from each group

  • Example:

    • Divide students in a class by major:

      • Marketing majors

      • Management majors

      • Accounting majors

    • Randomly select students from each major group to participate in the survey.


Cluster sample: the population is divided into groups (clusters), and then the researcher randomly selects certain clusters.

Picks certain groups and chooses everyone from the group




what are the different probability samples

39
New cards

Sampling Method

Cost/Ease

Statistical Results

Convenience

Very Low

Very Poor

Judgment

Low

Very Poor

Quota

Low

Very Poor

Simple Random (SRS)

Moderate

Fair

Systematic

Low–Moderate

Fair

Cluster

Moderate

Fair–Good

Stratified

High

Very Good


what are the pros and cons of sampling methods

40
New cards

data analysis: turning raw survey answers into useful information.

coding: Grouping and assigning values to survey data.

  • You need labels to understand what the assigned numbers mean.


  • dummy codes: Usually use 0 and 1.

    • Example:

      • College Student: No = 0

      • College Student: Yes = 1


effects codeUsually use +1 and -1.

  • Commonly used for experimental data.

  • Example:

    • College Student: No = -1

    • College Student: Yes = +1




what is basic data analysis and its codes


41
New cards

descriptive: Studying the basic characteristics and distribution of data.

  • Used to summarize and describe the data collected from a sample.

  • central tendency: Measures the center or typical value of a group of data.(mean, median, mode)


dispersion: Measures how spread out the values are around the mean.

range (Large-small)

standard deviation (Measures how far values typically are from the mean)

variance)

variance: measures the avg squared distance of values from the mean

what is the difference between descriptive statistics and dispersion

42
New cards


  • Tabulations: Organizing survey data into tables, charts, or figures to make results easier to understand.

  • Often makes data easier to interpret.

  • Helps identify errors or outliers (values that are very different from the rest).

  • Helps determine non-response on questions (when respondents leave questions unanswered).

  • ex) pie charts, bar charts, graphs)


what are tabulations/figures

43
New cards
  • Using results from a sample to make conclusions about the entire population.

  • Uses statistical tests to determine whether relationships or differences exist.


what are inferential statistics

44
New cards

1) Develop Hypothesis

  • Null Hypothesis (H₀): There is no difference or association between variables.

  • Alternative Hypothesis (H₁): There is a significant difference or association between variables.


2) Test Hypothesis

  • Select a statistical test to see if there is a relationship.

  • Look at the p-value to help determine whether the results are statistically significant.


3) Interpret Results

  • Reject the Null: There is enough evidence to support the alternative hypothesis. (there is enough data they are similar, goes against null hypothesis, kind of like admitting you were wrong)

  • Fail to Reject the Null: There is not enough evidence to support the alternative hypothesis.



what are the 3 steps of the inferential statistics

45
New cards

nominal/ordinal = categorial

  • categorial can not find the means

  • Categorical Data

    • Definition: Data that is separated into groups or categories.

    • Usually describes what type of something or which group someone belongs to.(gender, major, yes or no, fav restaurant)

    • What group?


interval/ratio= continuous

  • continuous can find the means

  • Continuous Data

    • Definition: Data measured using numbers that can have many possible values, including decimals.

    • Usually measures how much, how many, or how long (age, height, weight, income, time spent)

    • How much or how many?


what are the levels of scare measurement and suggested descriptive statistics

46
New cards

chi square test: Question: Are two variables related?

  • Uses 2 categorical variables (nominal or ordinal).

  • Example: Is college major related to gender?

Remember: Categorical + Categorical = Chi-Square


mean differences: Two types to figure out if means are different

1) Independent Samples T-Test

  • Uses 1 continuous variable (interval/ratio) and 1 categorical variable (nominal/ordinal).

  • Compares the averages of two separate groups.

  • Example: Do male and female students spend different average amounts on food?

2) Paired Samples T-Test

  • Uses 2 continuous measurements from the same participants or matched pairs.

  • Compares two related sets of scores.

  • Example: Comparing students' test scores before and after studying.


regression test: Question: How well does one variable predict an outcome?

  • Uses 2 continuous variables (interval/ratio).

  • Example: Does the number of hours studied predict exam scores?


multiple regression test: Question: How well do multiple variables predict an outcome?

  • Uses multiple continuous variables (interval/ratio).

  • Example: Do hours studied, class attendance, and sleep predict exam scores?


what are some statistical tests used

47
New cards
  • P-value smaller than 0.050 = Significant relationship/result.

  • P-value larger than 0.050 = Insignificant relationship/result.


significance level: A predetermined level set BEFORE running a statistical test.

  • Determines how strong the evidence needs to be to consider a result significant.

  • The most common significance level is 0.050 (5%).

  • If the p-value is larger than 0.050, the result is NOT significant.


p value: the calculated probability of getting results like the ones observed if the null hypothesis is true.

  • A p-value helps determine whether a result is statistically significant.

  • Example: A p-value of 0.020 is significant at a 0.050 significance level.


what are the basics of statistical tests

48
New cards

one way tabulation: Categorizing and counting responses for one variable.

  • Shows how many people fall into each category.

  • One-way = looking at ONE variable.


cross tabulation:Categorizing and comparing two or more variables.

  • Helps researchers see if variables are related.

  • Can calculate a chi-square statistic to determine if the relationship is statistically significant.

  • Uses 2 nominal- or ordinal-level (categorical) variables


what are the two data tabulations

49
New cards


For this test

The relationship between class rank and living on campus is statistically significant (χ² = 9.24, p-value = 0.002). Upperclassmen are more likely to live off campus, with 33 of 39 living off campus, while 8 of 14 underclassmen live on campus.



  • Chi-square = 9.24 (2 decimal places)

  • p-value = 0.002 (do not round)

  • Because p < 0.05, report the group values to explain the relationship


how do you report a chi square test

<p>how do you report a chi square test </p>
50
New cards

independent samples t test: Separate the data into two groups (male and female).

  1. Find the average exam score (mean) for each group.

  2. Compare the two group averages to see if they are significantly different.

Easy reminder: Two separate groups, comparing their averages.


paired samples t test: Find the average midterm score.

  1. Find the average final exam score.

  2. Compare the two averages to see if scores changed.

Easy reminder: Same people, measured twice.

how do you report mean comparison tests

51
New cards

1) identify your variables: categorial( which group) and continuous (how much or how many)

2) Know which test to use

Chi-Square Test

2 categorical variables

2 categories → Are they related? (Is gender related to college major?)

Independent Samples T-test

1 categorical + 1 continuous

2 different groups → Are their averages different?(Do men or women have higher exam scores?)

Paired Samples T-test

2 related continuous measurements

Same people, 2 measurements → Did their averages change? (Did students' scores improve from the midterm to the final?

how do you know what test to run

52
New cards

semantic differential scale because it measures a respondent's attitude toward a product, service, or concept using a series of bipolar (opposite) adjectives placed at either end of a multi-point continuum—typically 7 points

1. A grocery delivery company asks customers to rate its service using separate 7-point items

anchored by unreliable–reliable, slow–fast, and inconvenient–convenient. What type of scale is

the company using?

A. Graphic rating scale

B. Behavior intention scale

C. Semantic differential scale

D. Likert scale

E. Constant-sum scale

53
New cards

stratified random sampling because the researcher divides the entire population into distinct subgroups (called strata) based on a shared characteristic

A streaming service wants feedback from customers on its Basic, Standard, and Premium

plans. The researcher divides the customer list by plan and randomly selects customers from

each of the three groups. Which sampling method is this?

A. Simple random sampling

B. Systematic random sampling

C. Cluster sampling

D. Stratified random sampling

E. Quota sampling