Unit 5: Quantitative Techniques for Business
Questionnaires and Sampling
Click here to see the OHT slides from the lesson.
 |
Purpose of a Questionnaire
A questionnaire is simply a set of questions and its purpose is to obtain information.
It is an imposition and requires the goodwill of the respondent for its success.
Constructing Questionnaires
Object and Authority
- what is the enquiry trying to establish and why is it trying to establish it
- example: "This questionnaire is trying to establish the extent to which the recent flu virus and spread within the area. From the results it will be determined whether or not to offer the flu vaccine for free next winter"
Instructions
- How to Complete
- When to Complete
- How to return the Questionnaire
Details of Respondent
personal questions can be a sensitive area, restrict personal details
 |
The Body of the Enquiry
Don'ts
- multiple questions
- unfamiliar words
- multiple instructions
- calculations
- relying on memory
- inappropriate
- unnecessary
- tactless and offensive
- ambiguous
- biased questions
The Body of the Enquiry
- length of questions
- sequence of questions
- units to be used in answers
Surveys
- Pilot Survey
- Panel Survey
- Longitudinal Survey
- Market Research Survey
- Public Opinion Polls
Sampling
Random Samples
- the sample is taken from the population purely at random
- use random number tables or the random number feature on spreadsheet programmes (eg Excel)
Quota Samples
- the proportions of people to be chosen are pre-set
- the quota chosen is representative of the population in one or two respects but not necessarily in others
- for example: people with blue eyes
Multi-Stage Samples
- a random sample of a random sample
- for example: to get a random population from 100 streets in Bristol, randomly select 10 streets from the 100 and then select 10 households form the 10 streets
Cluster Sampling
- same as multi-stage sampling with the final sampling done according to a specific characteristic
- for example: taking our ten streets, we would ask ten men with dark hair
Systematic Sampling
- the sample is undertaken by choosing the respondents at regular intervals,
- for example: taking every tenth street from our list of 100 streets and then every tenth household.
- these samples are biased against those people who have no chance of been chosen once the first person has been chosen
Stratified Sampling
- the population are divided into blocks where all members have something in common, random samples are then chosen from these blocks
- for example: splitting the population into blocks according to age
