L8, Data collection 1 - quantitative Flashcards
Explain the differences between exploratory and confirmatory research in the early phases of the product development
?
Explain what is meant by primary research and exemplify the information sources that can be used
?
Relate different types of research methods/tools to different market research purposes
?
Plan and execute a study eliciting customer needs and requirements including purposeful choice between different methods and tools for eliciting, analyzing, communicating and documenting customer needs and requirements
?
Describe the principles of and apply the analysis tools for analyzing qualitative and quantitative data respectively
?
Why should we do a quantitative market research?
Because we don’t know:
- what customer actually need
- what’s on their mind or how their situation is changing
- what they are unhappy about
- what is driving happyness
- How many customers of different types there are out there?
- How customers select a vendor, search for information, decide where to shop
- What drives the choice of one product over another
- How much they are willing to pay
- How many buy at this price
Give examples of quantitative research!
1. Asking —> surveys 2. Observing —> big data 3. Arrange stimuli —> experiments —> choice modeling
What is a survey?
- Survey research entails the collection of data on a number of units
- Usually at a single juncture in time, with a view to collect systematically a body of quantifiable data in respect of a number of variables
THEN examined to discern patterns of association
Central features of a survey
- the use if a FIXED set of questions directed to a large set of respondents
- The collection of a small amount of data in standardized and quantified form from a relatively large number of individuals
- The selection of representative samples of individuals from known populations
Describe the survey research process
- planning
- Questionnaire design
- Decide on sample size and source
- Pretest survey
- administer survey
- Survey data analysis
What should be done during the survey planning?
- state purpose of the study
- Identify kinds of information of interest
- Identify population to be surveyed
- Indicate sample size and source
Give examples of purpose of survey!
- Understand how satisfied our customers are
- Determine sizes of market segments
- Understand product usage and ownership
- estimate purchase intention during next year
- Evaluate and compare brand image and perception
- Tracking changes in customer demographics and product usage over time
What information should be interesting and should be include in the survey (examples)
- user demographics
- usage patterns
- enchanted problems
- ability to pay
- knowledge of own and competitor brand
How do you identify population to be surveyed?
SAMPLING
A good sample is essential for the generalizability of the results, but it is very hard to get
What is a sampling population?
The total set of people that are related to a particular investgation
What is a sample?
A selection from the population
What is a sample frame?
The source from which the selection is made —> the payroll of a company —> the phone directory —> the national census —> the car brand club members
what different sample types are there?
- Probability samples
- Non- probability samples
- Census
What is a probability sample?
The probability for selection of each individual is known
Statistical methods assume probability samples
— simple random sampling — systematic sampling — stratified random sampling — cluster sampling — multistage sampling
— large number of respondents (+100)
what is a non-probability sample?
The researches uses the judgement to select respondents suitable for the purpose
— quota samling — dimensional sampling — convenience sampling — purposive sampling — snowball sampling
What is a census sample?
The whole population is surveyed
Financially viable when the population is very small
Describe the process of questionnaire design!
- Generate list of potential topics and kind of information to be collected
- Construct a first draft
—> including actual questions and answer categories
—> re-use own questions or from existing questionnaires - Put it aside for a few days, then revise
- Circulate for the comments and critique
- produce another draft
- circulate comments
- produce final draft
- pre-test
- produce final version
Advice for questionnaire in general
- maximize the reward and minimize the costs for responding
- shorter is better
- cluster contents and seek natural flow
- place demographics and classification data last
- starting with relevant & interesting questions
- professional appearance
Advice for question phrasing in the questionnaire
- keep it simple
- be specific
- ask mostly closed-ended questions
- always ask some open-ended questions
- mnimize demands on memory
- match questions to how the market works
- avoid loaded questions
Advice regarding answer categories in questionnaires?
- use the right number of answer of answer categories
—> yes-no is sometimes enough
2.distinguish between bipolar and unipolar responses
—> BIPOLAR: qualitative answer categories
—> UNIPOLAR: quantitative from 0-100% e.g.
- attend to the interaction of the answer categories
- Beware of including don’t know and not applicable
- prefer ratings over rankings
What different types of question are there?
- categorization
B2C: age, gender, income, education, current brand
B2B: employees, turnover, role in company - Descriptive questions
—> How often do you use our product
—>what is the frequency of product X - Rating scales
—> how satisfied are you with our product’s performance
Create a small questioner to understand customer needs of a industrial forklift!
DEMOGRAPHICS:
1. what is the turnover of your company
< 10 MEUR, 10-100 MEUR, >100 MEUR>
- How many employees do you have?
< 100, 100-500, 500-500, >5000>
DESCRIPTIVE:
3. What is the average daily driving distance?
- Maximum shelf height
5 current brand
RATING:
The ergonomics of forklifts is not affecting pain in the body
< strongly disagree, disagree, agree, strongly agree>
It doesn’t matter if we have different types of brands of the fork lifter
—-
How do you decide on sample size and source?
- what levels of confidence and precision do you want from your study
- how many respondents are needed to reach desired levels of confidence and precision?
- Given number of responses, what can be claimed about confidence level and precision?
Formula for computing sample size N?
N=Z^2*variance/ precision^2
Z= confidence level= how likely would be to get the same difference by chance alone if there really is no difference in the population between categories represented by your group
Variance
= the variance in the population of the quantity to be estimated
= the average of the squared deviations of individual scores from the mean
Precision= the decision maker’s tolerable margin of error (compare with confidence interval)
Name some tools for surveys!
- SurveyMonkey
- SurveyGizmo
- SAS
- SPSS
- Mentimeter
Name some different Administer surveys
- postal surveys
- telephone interviews
- video conference interviews
- Face-to-face interviews
- Web-based survey
Give example of survey data analysis
- Create a data set e.g. in an excel worksheet or in a dedicated statistical analysis software
- clean up the data set
- Analyze basic metrics and basic diagrams
- Analyze relationships between two variables
- Analyze relationships between three or more variables
Name some few basic diagrams
- pie chart
- histogram
- tables
- box plot
- cumulative plot
- regression, distribution
Strengths and weaknesses of questionnaire-based surveys
STRENGTH
1. Precise numerical estimates of the frequency and magnitude of consumer responses
—> useful when precision matters a lot
- Objective
- reaches many people
- avoids personal prejudice and limitations
- allows to statistical methods
- identifies and illuminates differences between groups
- can be repeated over time
WEAKNESSES
- Tells you what but not why
- cannot tell you what you didn’t know that you didn’t know
- rely on self-report
- participation is unrewarding
- surveys are only as good as the questions and the sample
- sometimes decision-making is done by groups
Conclusions with regards to surveys?
- easy and cheap to create and distribute
- analysis of survey data:
—> identifies important numbers
—> enables analysis of related aspects
—> enables evolution of quantities over time - Surveys are an important tool for:
—> product planners
—> guiding improvement of existing products
—> less applicable on completely new products - Statistical significance of findings requires
—> good questionnaire design
—> large number of properly sampled responses - Surveys are mainly used for confirmatory research
—> could to some extent be used e.g. through inclusion of open-ended questions
