Research Methodology

1
Research Methodology

• Lecture No 16
• ( Sampling / Non Probability, Confidence and
  Precision, Sample size)

2
Recap Lecture

• Systematic ,stratified sampling, cluster, area
  and double sampling are the common types of
  complex sampling.
• Convenience, judgment, quota and snowball
  sampling are the common types of non probability
  sampling.

3
Lecture Objectives

• Non Probability Based sampling (Quota/snow ball)
• Discuss about the precision and the confidence.
• Precision and Confidence
• Factors to be taken into consideration for
  determining sample size.
• Managerial implications of sampling.

4
Non-Probability Sampling

• Quota Sampling
• This is a sampling technique in which the
  business researcher ensures that certain
  characteristics of a population are represented
  in the sample to an extent which is he or she
  desires.

5
Non-Probability Sampling

• Quota Sampling
• Example A business researcher wants to determine
  through interview, the demand for Product X in a
  district which is very diverse in terms of its
  ethnic composition.
• If the sample size is to consist of 100 units,
  the number of individuals from each ethnic group
  interviewed should correspond to the groups
  percentage composition of the total population of
  that district.

6
Quota Sampling
Example Quotas have been set for gender only.
Under the circumstances, its no surprise that
the sample is representative of the population
only in terms of gender, not in terms of race.
Interviewers are only human.
7
Non-Probability Sampling

• Snowball Sampling
• This is a sampling technique in which individuals
  or organizations are selected first by
  probability methods, and then additional
  respondents are identified based on information
  provided by the first group of respondents

8
Non-Probability Sampling

• Snowball Sampling
• The advantage of snowball sampling is that
  smaller sample sizes and costs are necessary a
  major disadvantage is that the second group of
  respondents suggested by the first group may be
  very similar and not representative of the
  population with that characteristic.
• Example Through a sample of 500 individuals, 20
  antique car enthusiasts are identified which, in
  turn, identify a number of other antique car
  enthuiasts

9
More Snowball Sampling
More systematic versions of snowball sampling can
reduce the potential for bias. For example,
respondent-driven sampling gives financial
incentives to respondents to recruit peers.
10
Issues in Sample Design and Selection

• Availability of Information Often information
  on potential sample participants in the form of
  lists, directories etc. is unavailable
  (especially in developing countries) which makes
  some sampling techniques (e.g. systematic
  sampling) impossible to undertake.

11

• Resources Time, money and individual or
  institutional capacity are very important
  considerations due to the limitation on them.
  Often, these resources must be traded against
  accuracy.

12
Issues in Sample Design and Selection

• Geographical Considerations The number and
  dispersion of population elements may determine
  the sampling technique used (e.g. cluster
  sampling).
• Statistical Analysis This should be performed
  only on samples which have been created through
  probability sampling (i.e. not probability
  sampling).
• Accuracy Samples should be representative of
  the target population (less accuracy is required
  for exploratory research than for conclusive
  research projects).

13
Issues of precision and confidence in determining
sample size

• Precision
• Precision is how close our estimate is to the
  true population characteristic.
• Precision is the function of the range of
  variability in the sampling distribution of the
  sample mean.

14
Population and Sample distinctiveness

• Sample Statistics( Mean, Std Deviation, Variance)
  and Population parameters ( Mean, Std Deviation,
  Variance)
• Compare the Sample estimates and population
  characteristic. Where the estimates should be the
  representative of the population charactertics
• Sample statistics (mean, sd, ..) should be
  representative of the population parameters(mean,
  sd )

15
Issues of precision and confidence in determining
sample size

• Precision
• How close are the estimates to the population.
• While expecting that the population mean would it
  fall between (,- )10 points or (,-) 5 points
  based on the sample estimates is precision.
• The narrower the more precise our statement is

16

• E.g The average age of the a particular class
  based on the sample is between 20 and 25
• Or it between 18 and 28.
• How close are the estimates to the population.

17

• Confidence
• Confidence denotes how certain we are that our
  estimate will hold true for the population.
• The level of confidence can range from 0 to 100.
  However 95 confidence is the conventionally
  accepted for most business research.

18

• The more we want to be precise the less confident
  we become that our statement is going to be true.
• So at one level we want to be accurate in our
  statement but on the other we taking a higher
  risk of proved incorrect.
• In order to maintain the precision and increase
  the confidence or increase the precision and the
  confidence we need to have a larger sample.

19
Determining sample size

• Roscoe (1975) proposes the following rules of
  thumb for determining sample size.
• Sample sizes larger than 30 and less than 500 are
  appropriate for most research
• Where sample sizes are broken into subsamples
  (males/females, juniors/seniors etc.), a minimum
  sample size of 30 for each category is necessary.

20
Determining sample size

• In multivariate research (including multiple
  regression analysis), the sample size should be
  several times (preferably ten times or more) as
  large as the number of variables in the study.
• For simple experimental research with tight
  experimental controls (matched pairs, etc.),
  successful research is possible with samples as
  small as 10 to 20 in size.

21

• Tools and mathematical equations are available to
  establish the right size of the sample.
• Refer to the book for the sample size calculation
  equation.
• Standard Tables are available
• Use a software like RAO calculator available on
  the internet.

22
Types of Sampling Designs
23
Managerial Implications

• Awareness of sampling designs and sample size
  helps managers to understand why a particular of
  sampling is used by researchers.
• It also facilitates understanding of the cost
  implications of different designs, and the trade
  off between precision and confidence vis-à-vis
  the costs.

24
Managerial Implications

• This enables managers to understand the risk they
  take in implementing changes based on the results
  of the research study.
• By reading journal articles, this knowledge also
  helps managers to assess the generazibility of
  the findings and analyze the implications of
  trying out the recommendations made therein in
  their own system.

25
Recap

• Non Probability based sampling (
• Precision we estimate the population parameter to
  fall within a range, based on sample estimate.
• Confidence is the certainty that our estimate
  will hold true for the population.
• Roscoe (1975) rules of thumb for determining
  sample size.
• Some sampling designs are more efficient than the
  others.
• The knowledge about sampling is used for
  different managerial implications.