L3 Quantitative Research Flashcards
What is Quantitative Research
Confirmatory
- reality is objective and identifiable (behavioral psychology, economics, statistics)
- surveys, experiments, conjoint
Conjoint analysis
is a survey-based statistical technique used in market research that helps determine how people value different attributes (feature, function, benefits) that make up an individual product or service.
example descriptive questions:
e.g.: - consumer characteristics: what percentage of consumers are in their 30s?
What is a good descriptive research?
minimizes the error in representing reality
What types of error do we have?
Total survey error = Random error + Systematic error (bias)
random error
difference btw. the sample value and the true value of the population mean (no direct control over this, results form chance variation)
systematic error
- problems in the very research design (measurement error )
- problems in the sampling process (sample design error)
two types of systematic error
measurement error and sample design error
scales
sets of symbols or numbers so constructed that the symbols or numbers can be assigned by a rule to the individuals (or their behaviors or attitudes) to whom the scale is applied
- this is about how you can treat your data (not objective)
interval
no true zero
mode, median, mean
ordinal
(mode, median) relative preferences (objects are greater or smaller)
ratio
there now is a real zero
How to assign numbers or symbols?
- In a way that the latent construct can be observed,
- information reflects what is needed
Most popular scaling techniques
comparative scaling
and non-comparative scaling
comparative scaling (most popular techniques)
- paired comparison scaling
- rank order scaling
- constant sum scale
non-comparative scaling (most popular techniques)
- likert scale
- semantic differential scale
Issues with non-comparative scales
- how many points on the scale? (Richness: not too few vs. discriminability: not too many=
- balanced or unbalanced (more options for left side or right side? vs. see options on both sides): use balanced, unless distribution is very skewed
- odd- or even-numbering scales (do you want to force preference?)
- force choice..
reliability
degree to which measures are free from random error and, therefore, provide consistent data. The extent to which the survey responses are internally consistent
How to test for reliability?
- test and retest
- equivalent form (do highly similar tools produce the same result?)
- internal consistency (do different parts of the tool produce the same results?) (Cronbach’s alpha)
reliability vs. validity
Validity= are the measures appropriate and meaningful? Reliability= Possible to repeat or corroborate the findings?
corroborate
confirm or give support to
how is zero interpreted
absence of success
-5 interpreted
presence of failure
Sampling methods
probability sampling
non-probability sampling
probability sampling
systematic, stratified, cluster, simple random
non probability sampling
convenience, snowball, judgment, quota