Market Research (3) Flashcards
(22 cards)
How many study design types are there in the field of Market research?
In the field of Market research there are:
1) Exploratory : large topic, in-depth interviews
2) Descriptive: specific focus and surveys
3) Causal: identification of mechanisms
Where can I execute the different advertisement research?
1) Internally Executed (for privacy)
2) Externally Executed ( need methodological knowledge)
How many different methods are there to select and sample participants?
1) Simple Random Sample
2) Stratified Random Sample: Divide into layers and randomly select while accounting for representativity.
What are metric and non metric data?
metric = quantitative
non-metric = non quantitative
Which like/dislike scales are used for experiments in the realm of marketing?
Usually with children they use graphical scales (Lawless) or the so-called Snoopy scale by Scharf.
What kind of Datatypes can be collected?
Data can be primary or secondary depending on the source and the precision of this data.
Primary data= Collected in a specific manner by the author.
Secondary data= from periodical sources like Statista or Google Trends.
What are the benefits and the disadvantages of primary data in comparison to secondary data?
Primary data has a more IN-DEPTH view but is also HIGHER in costs and more time demanding.
Secondary data has lower costs but also a BROADER view.
What are the differences and the implication of Laboratory and Field experiments in marketing?
Laboratory experiments: internal validity better control of variables but might undergo weird effects for the UNKNOW ENVIRONMENT.
While Field experiments are also MORE EXPENSIVE but more realistic observations (low internal validity).
If I have a causal research question in marketing which type of study should I use?
For causality research, you should look into experimentation of eg. advertising campaigns & AB testing.
What is a stratified random sample?
A stratified random sample divides the population into different categories and ensures to take a random sample from each one of these. The goal is to have a representative cluster.
What is ANOVA? What is its purpose?
ANOVA is the abbreviation of Analysis of Variance. It calculates the difference between different groups while still accounting for the normal variation within each group.
What does an ANOVA result indicates?
An ANOVA result indicates whether there is a statistical significance in the occurance of a result as part of a marketing research experiment.
What is the GENERAL LINEAR MODEL in ANOVA? What is it composed of?
y= u + a + e
u= population mean
a= effect under independent variable
e= error
Can you explain what is SStotal, SSbg and SSwg?
SSbg = sum of squared deviations (within each group mean and the GM
SSwg= sum of squared deviation between each score and group mean.
SStotal= total sum of squared deviations between scores and GM (of the scores).
What is MSbg and MSwg?
MSbg= total variability between groups (how much are they different from each other?)
MSwg= total variability of the scores within a group (how much can a result vary but still be part of this category/group?)
What is the F value? What does it describe? What would a -1 score indicate?
The F values rappresents the difference between the variance between groups vs the variance within the groups. A -1 score indicates that there is not much differnece between or within the scores = no statistical relevance»> our hypothesis is null.
What is the difference between Descriptive, Predictive and Prescriptive statistics?
Descriptive = describe things of the past
Predictive = describe things of the future
Prescriptive = suggest a sequence of norms to obtain a specific result. Suggest a COURSE OF ACTION.
Offer me an example of A/B testing?
Assess whether a blue banner o a red banner might work better. Assess whether a big button or a smaller button might bring more profit. Measurable feature? Profit gains.
What is Silicon Sampling?
Silicon sampling uses LLM (Large language Models) to produce syntetic users that can be explored by researchers to improve their product.
Why is Silicon sampling frequently used and what are some implications?
Silicon Sampling is frequently used for their cost-efficiency. However, how good are these synthetic users accurate when it comes to SENSORY experiences? They have reached negative results when considering marketing suggestions for CONGRUENT enviroment stimuli.
What are the barriers of the attitude-behaviour gap?
1) Ineffectiveness assumption ( I will not make a difference)
2) Self-interest orientation (profit? then no)
3) Mistrust (in companies or in people)
What is the critical F and how does my own F has to be in order for us to have statistical significance?
Our critical F indicates the minimum for it to be statistically significant. This mean that our F value has to be bigger then the critical one.
