Help
Options
Didn't know it?
click below
click below
Knew it?
click below
click below
Don't Know
Remaining cards (0)
Know
retry
shuffle
restart
0:00
Embed Code - If you would like this activity on your web page, copy the script below and paste it into your web page.
Normal Size Small Size show me how
Normal Size Small Size show me how
Marketing research
Lecture 2
| Question | Answer |
|---|---|
| What are the advantages of multiple choice questions? | It’s easy to register, reliable and easier to code/interpretation, analyse. Open questions have a lot of categorisation afterwards. |
| What are the disadvantages of multiple choice? | Limited answering options and respondent is helped too much |
| How do you prevent biases? | Through wording, you have to formally questions carefully to improve data quality, avoid leading questions. Avoid vague words, instead ask actual frequency. |
| What else should you avoid when formulating questions? | Avoid double-barrelled questions and avoid pies for ordinal scales |
| What are scales? | The assignment of numbers, the characteristics of objects, according to pre-specified rules |
| What are the four main scale types for measurement? | Nominal- mode Ordinal- mode plus median Interval- mode plus median plus mean Ratio- all-stats |
| What can you have respondents do ? | Compare objects or rate independently |
| Comparative | Scale objects relative to other objects, options include paired comparison, rank order and constant sum |
| Non-comparative (metric) | Scale independently of other objects, options include continuous scale, likert and semantic differential (bipolar) |
| What decisions do you have to make when making a survey? | Number of scale categories, odd versus even number of categories, balanced versus unbalanced scale |
| When do you use I don’t know option in the survey? | Only in specific cases, use when requesting factual information, but not when requesting opinions |
| Reliability and validity of the measurement | Reliability: if I repeat the same measurement, do I get the same result? Validity: am I measuring what I want to measure? Similar measurement techniques should give the same result, often used face validity. |
| Reliability: repeated measurements give same results | Test and retest, reliability, internal consistency. Reliability can be checked for multi item scales. How do you compute this? through Cronbach alpha or factor analysis |
| Validity: do the measurement instruments do what they’re supposed to | Face validity is often used. It tells us if all aspects are covered in the measurement and it’s relatively subjective. Criterion validity is more formal. Bad surveys can lead to invalidity. |
| How do you improve validity in surveys? | By using validated skills, let experts judge survey questions, and trying it out on a few individuals |
| Reliability versus validity | Hard to prove validity, but reliability is a necessary condition, however, not sufficient |
| What is dimension reduction? | Combining items into fewer variables using cronbachs alpha and factor analysis |
| When+Why do you combine variables? | - To check internal consistency - To have a 'backup' variable - Usually done with >2 variables - when they correlate enough you can combine into 1 construct |
| Cronbach's Alpha | CA depends on umber of variables , correlations and total variance. Higher correlations result in higher CA . You want internal consistency as high as possible in order to be allowed to combine |
| Discovering which items you can combine with factor analysis | Use principle component analysis to make the split for you. Variables can be expressed as a linear combination of other variables called factors. number of components unknown, linear combo is expressed with loadings |
| What are the steps for factor analysis? | - research purpose -compute PCA solution - check if FA is appropriate - select number of factors - Rotate factor matrix with loadings - interpret factors based on loadings |
| What does rotation do? | Rotation makes loadings more extreme so that items belong more strongly to a component -- easier interpretation |
| When do you use exploratory research? | to decide which attributes need to be measured. Use pre-stage qualitative research, and determine two things for the attributes found; importance and performance |
| Performance | Measure overall and separate per component |
| Importance | Ask directly or derive from importance scores - Stated importance: direct questioning - Derived importance: from performance grades using linear regression |
| What is relative importance derived from? | Co efficients |
| Net promoter scale | Measurement of loyalty, only 1 question Groups: Promoters, passives and detractors NPS= %Promotes - % detractors |
| What are the drawbacks of NPS? | Just 1 question, no check for internal consistency possible, response style has an influence. 9-10 mark is high mark to achieve |
Created by:
janasultan
Popular Marketing sets