Term
strong positive relationship
|
|
Definition
|
|
Term
| strong positive relationship correlation |
|
Definition
|
|
Term
| Weak Positive Relationship |
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
| marketing example of strong positive relationship |
|
Definition
|
|
Term
| marketing example of weak relationship |
|
Definition
| price and quality [ex. Jaguar car...price is high and doesn't have as high of quality (as in reliability) compared to the price] |
|
|
Term
| marketing example of strong negative relationship |
|
Definition
|
|
Term
| marketing example of no relationship |
|
Definition
| satisfaction level of In-N-Out vs. satisfaction level of Nordstroms |
|
|
Term
| marketing example of weak positive relationship |
|
Definition
|
|
Term
|
Definition
Finds relationships
all respondents interpret the question the same
-more items means more reliability
-new construct variables are usually more reliable than single items measured |
|
|
Term
|
Definition
Are we measuring what we say we are measuring?
-Concept
-Face |
|
|
Term
|
Definition
you don't use price to measure commitment...you don't use satisfaction to measure trust
Are we measuring the right thing?? |
|
|
Term
|
Definition
if Hyundai does better than BMW, something's amiss
does the outcome make sense?? |
|
|
Term
|
Definition
we want both!
independent of each other |
|
|
Term
| more factors (Questions about the same thing) equals... |
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
| bell curve of reliable info |
|
Definition
|
|
Term
| Bell Curve that is Skewed / Biased |
|
Definition
|
|
Term
| Importance vs. Significance |
|
Definition
| Both matter, but IMPORTANCE matters more |
|
|
Term
| importance without significance... |
|
Definition
|
|
Term
| significance without importance... |
|
Definition
|
|
Term
| significance driven, in large part, by ___ |
|
Definition
|
|
Term
| small correlation with high sample size = |
|
Definition
|
|
Term
| large correlation with small sample = |
|
Definition
|
|
Term
| Disregard a sample size below... |
|
Definition
|
|
Term
| What drives statistical significance? |
|
Definition
|
|
Term
| Look at two graphs on importance vs. significance page! |
|
Definition
|
|
Term
| What is multicollinearity? |
|
Definition
| 2 independent variables that are highly related |
|
|
Term
| How does multicollinearity affect Regression? |
|
Definition
| it can't rank variables properly |
|
|
Term
| How do you deal with multicollinearity? |
|
Definition
-Reduce items...Factor Analysis (forms constructs)
-Use Correlation to rank predictors |
|
|
Term
| Why is regression so often misused? |
|
Definition
-over/under estimates impact
-wrong ranking |
|
|
Term
|
Definition
|
|
Term
| Difference between Causality and Correlation |
|
Definition
-relationship &
-temporal order |
|
|
Term
|
Definition
| one has to come after the other |
|
|
Term
| when doing survey research... |
|
Definition
| ...we hypothesize a "causal picture" aka Model and test it |
|
|
Term
|
Definition
|
|
Term
| The only way to prove causality |
|
Definition
|
|
Term
|
Definition
|
|
Term
| causality examples from class |
|
Definition
|
|
Term
| Come up with a causality example with a lurking variable!! |
|
Definition
|
|
Term
| If you have issues with multicollinearity what will you do to fix it? |
|
Definition
-Factor Analysis (gets rid of redundancy)
- Correlations (ranks predicters) |
|
|
Term
| Whats the deal with surveys and multicollinearity? |
|
Definition
Surveys = lots of multicollinearity
2 questions on survey that are next to each other = automatically correlated (illusion at least) |
|
|
Term
| Which perf predictors will suffer the most from multicollinearity? |
|
Definition
| Run a correlation and the 2 predictors with the highest number = the ones that will suffer the most |
|
|
Term
| Why do we care about reliability? |
|
Definition
all respondents interpret the question the same = reliability
It helps us determine relationships actually exist |
|
|
Term
| What happens if we don't have reliable measures? |
|
Definition
|
|
Term
| Whats the relationship between reliability and validity? |
|
Definition
| independent, but we want both |
|
|
Term
| When we're doing factor analysis, where does the concept of validity fit in? |
|
Definition
In the naming process.
You want to make sure that the items that make up the construct accurately represent the construct |
|
|
Term
| What do we do if we have a significant, but unimportant relationship? |
|
Definition
move on, don't do anything with it
don't base any decisions by it |
|
|
Term
| How do we increase significance without increasing importance? |
|
Definition
|
|
Term
| What is a lurking variable? |
|
Definition
| a variable that really predicts two outcomes but tends to be forgotten so we think that one factor predicts another when really the lurking variable predicts both |
|
|
Term
| What do I mean when I suggest that relationships have Personal and Functional Dimensions? |
|
Definition
| Personal and Functional Dimensions contribute to Commitment! |
|
|
Term
| How do Trust and Satisfaction contribute differently to Personal and Functional connections? |
|
Definition
Trust- doesn't just satisfy you, but the company/service won't screw you over either
Satisfaction- meeting expectations |
|
|
Term
| What is the #1 thing to do? |
|
Definition
|
|
Term
|
Definition
| turning qualitative data into quantitative data |
|
|
Term
|
Definition
relationship between categorical and continuous variables
categorical variable = independent variable!
looks at overall differences…or average differences |
|
|
Term
|
Definition
the relationship between two Categorical
tell us is ‘level’ matters…e.g., does it have to be ‘top box’ to make a difference? |
|
|
Term
| If we want to know the % of Males/Females in data set we run… |
|
Definition
|
|
Term
What is coding…why do we do it?
What do we do when we code? |
|
Definition
Coding = turning qualitative data into quantitative
turning survey responses into analyzable data
reducing the number or response categories (age, income, etc.)
assigning numbers to all survey responses (very satisfied = #) |
|
|
Term
| Research techniques should follow... |
|
Definition
|
|
Term
| The key to successful marketing research... |
|
Definition
| is the identification of an important question/problem and maintaining research problem integrity throughout the data collection and analysis process |
|
|
Term
| There are tons of errors and many traps when analyzing survey data. We find the truth behind the numbers by surrounding it with ... |
|
Definition
| Triangulation-- answering a question using multiple methods |
|
|
Term
|
Definition
2 continuous variables
allows us to quickly compare strength of relationship…you can look at a lot of relationships at once. |
|
|
Term
|
Definition
| multiple continuous, independent variables and one dependent variable |
|
|
Term
| What alerts us to data problems? |
|
Definition
-Low variance (standard deviation)…low variance makes it hard to find relationships, even if they exist.
-Missing data…some analyses (regression, mean comparison etc.) won’t take respondents with missing data
-Reverse coding issues…when low is high and high is low |
|
|
Term
| When you reduce number of categories... |
|
Definition
| ...recode into different variable |
|
|
Term
| When you calculate new variables use... |
|
Definition
|
|
Term
| When you get rid of missing data... |
|
Definition
| ...recode into same variable |
|
|
Term
| What is the relationship between research question and research technique? |
|
Definition
| research techniques should follow research questions |
|
|
Term
| What are we trying to accomplish in Pre-Analysis |
|
Definition
-Identify key data characteristics, averages, variances ranges etc.
-Alert us to data problems
-Recode where necessary |
|
|
Term
| Maintaining research problem integrity is difficult…because: |
|
Definition
-Everybody wants a piece of the action
-The customer disappears
-Research takes time |
|
|
Term
| Nearly every survey you see in your marketing career... |
|
Definition
... will be too long
=respondent fatigue! (can loose focus) |
|
|
Term
| When do we use correlation? |
|
Definition
-Rank predictors
-Look for strong relationship among a large group of variables
-Alerting us to potential problems with Multicollinearity |
|
|
Term
|
Definition
-Turning one variable into a different variable…”transforming” variables
Recode into the same variable
Recode into a different variable
Compute a new variable |
|
|
Term
| Recoding into the same variable |
|
Definition
reverse coding
replacing the mean
get rid of missing data |
|
|
Term
| Recoding into a different variable |
|
Definition
-turning a continuous variable into a categorical variable
-collapsing categories |
|
|
Term
|
Definition
-building a construct
-creating groups or segments |
|
|
Term
|
Definition
a variable made up of other variables
ex. averaging many questions into one question
transform -> compute |
|
|
Term
| Why are there boxes and circles on the regression process model? |
|
Definition
Circles = constructs
squares = individual variables |
|
|
Term
| How did we determine the relative impact of trust and satisfaction on Loyal Behavior? |
|
Definition
| add both numbers up, then divide |
|
|
Term
| With factor analysis we can answer... |
|
Definition
Which items group together into constructs
How many constructs to we have
it says which variables are similar and helps get rid of redundancy! |
|
|
Term
| We do ____ before we compute new variables! |
|
Definition
|
|
Term
| The art (judgment) and science (numbers) of marketing analysis is... |
|
Definition
|
|
Term
| When running a factor analysis, don't forget to... |
|
Definition
Extraction > select Scree plot
Rotation > Varimax
Options >
Sorted by Size
Suppress absolute values less than… .50 |
|
|
Term
| Interpreting a Prescriptive Quad Map |
|
Definition
|
|
Term
|
Definition
|
|
Term
| Which picture: Describes how the world works? |
|
Definition
|
|
Term
| Which picture: Uses mean scores to map strength? |
|
Definition
|
|
Term
| Which picture: shows our brand only? |
|
Definition
|
|
Term
| Which picture: Maps the competitive landscape? |
|
Definition
|
|
Term
| Which picture: uses gap scores? |
|
Definition
|
|
Term
| Which picture: tells us what to do? |
|
Definition
|
|
Term
| Which picture: uses regression? |
|
Definition
|
|
Term
| How do we determine the variables most susceptible to multicollinearity? |
|
Definition
|
|
Term
| How do groups differ on key variables? |
|
Definition
| Cross Tabs and Mean Comparison |
|
|
Term
| Do individual items fall into related groups? |
|
Definition
|
|
Term
| Do we have redundant items in the survey? |
|
Definition
| Factor Analysis or a Correlation |
|
|
Term
| How much of the world is explained by our survey? |
|
Definition
|
|
Term
| Which items are the most important predictors? |
|
Definition
|
|
Term
| Do we have a relationship? |
|
Definition
| Correlation, Cross Tabs, Mean Comparison, Regression |
|
|
Term
| The more complicated and profound the findings... |
|
Definition
| ...the more linear and graphic the presentation must be |
|
|
Term
| Triangulation applies to... |
|
Definition
| ...presentations as well as it does to data analysis |
|
|
Term
| The fewer points you try to make... |
|
Definition
| ...the more likely they will each be recalled and used |
|
|
Term
|
Definition
Match the proportion in your data to proportion in the population…on an important attribute
-Variance...
-Salience |
|
|
Term
|
Definition
-Identify the proportion in the population
-Find the proportion in the data
-Multiply either variable by “weighting factor” to match proportion |
|
|
Term
| How do we know how many factors we have among image variables |
|
Definition
-Eigenvalue criteria
-Scree plot
-Eyeball test |
|
|
Term
|
Definition
| The difference between our score and the average of our competitive set |
|
|
Term
| What does the mean comparison tell us? |
|
Definition
overall differences
average differences |
|
|
Term
| What does correlation tell us? |
|
Definition
| the strength of a relationship |
|
|
Term
| What does a relationship map show us? |
|
Definition
identifies strengths and what kind of relationship you have (nature of relationship)
The number in the bubble is the commitment index |
|
|
Term
|
Definition
| used to compare different brands based on their group scores |
|
|
Term
| Regression is used to find... |
|
Definition
|
|
Term
|
Definition
the variable that we typically manipulate
aka the predictor |
|
|
Term
|
Definition
| the outcome, what we want to happen, the thing that changes based on the change to the IV |
|
|
Term
| What do we create when we code? |
|
Definition
|
|
Term
| Commitment is made up of _____ and ____ relationship |
|
Definition
|
|
Term
|
Definition
| standard deviation used to find a relationship |
|
|
Term
| What does cross tabs tell us? |
|
Definition
|
|
