Introduction to Sampling

1
Introduction to Sampling

• Mari Jack

2
(No Transcript)
3
Types of Sampling

• Probability Sampling
• Nonprobability Sampling

4
Probability Sampling

• Simple Random Sampling

5
Probability Sampling

• Stratified Random Sampling

6
Probability Sampling

• Systematic Random Sampling

7
Probability Sampling

• Cluster (area)
• Random
• Sampling

8
Probability Sampling

• Multistage Sampling

9
Nonprobability Sampling

• Accidental/haphazard or convenience sampling

10
Nonprobability Sampling

• Purposive Sampling
• Modal instance

11
Nonprobability Sampling

• Purposive Sampling
• Expert Sampling

12
Nonprobability Sampling

• Purposive Sampling
• Quota Sampling
• Proportional
• Nonproportional

13
Nonprobability Sampling

• Purposive Sampling
• Heterogeneity Sampling

14
Nonprobability Sampling

• Purposive Sampling
• Snowball Sampling

15
References

• Babbie, Earl. The Practice of Social Research.
  Belmont Wadsworth Cengage Learning, 2009.
• Kerlinger, Fred N. Foundations of Behavioral
  Research. Fort Worth Harcourt Brace Jovanovich
  College Publishers, 1986.
• Nachmias, David and Chava Nachmias. Research
  Methods in the Social Sciences. New York St.
  Martins Press, 1976.
• Rosenthal, Robert and Ralph L. Rosnow. Essentials
  of Behavioral Research Methods and Data
  Analysis. New York McGraw Hill Inc., 1991.
• Trochim, William M. K. Sampling
  http//www.socialresearchmethods.net/kb/sampling.p
  hp Research Methods Knowledge Base (accessed
  November 11, 2009).

16
How big should your sample be?

• Rule of thumb Bigger is better
• As n -gt N, Confidence Interval -gt 0

17
SE s/vn
18
A handy graph
19
Or is a chart better?
20
But is it really that easy?

• All equations for n given N are based on
  parameters of the population such as the standard
  deviation
• Kaplans paradox of sampling
• On the one hand, the sample is of no use if it is
  not truly representative of its population, if it
  is not a fair sample. On the other hand, to
  know that it is representative, we must know what
  the characteristics of the population are, so
  that we can judge whether the sample reflects
  them properly but in that case, we have no need
  of the sample at all.
• (Qtd. in Rosenthal 261)

21
Considerations for choosing sample size

• The degree of precision required between the
  sample and population
• Less precision smaller sample
• variability of the population
• A more homogenous population requires a smaller
  sample
• method of sampling
• A stratified sample requires fewer cases for
  accuracy
• way in which results will be analyzed
• A smaller sample puts limits on types of analyses
  possible

(Powell 105-6)
22
Statistical Power

• Power 1 - Beta
• Beta is the chance of error in rejecting the
  alternative hypothesis when it is true or
  accepting the null hypothesis when it is false
• ideally power should be .99
• usually settle for between .7 and .9

(Kraemer 10)
23
Considerations for choosing acceptable Power

• Specification of research goals in precise and
  realistic terms
• Identification of the design and measurement
  options available to address the research
  question
• Evaluation of the resources (time, personnel, and
  funding) available to the project.

(Kraemer 17)
24
How Power Effects Results

• Power dictates Critical Effect Size
• a measure of how strong the theory must minimally
  be to be important to society
• Can vary by field
• Sometimes a case study of n1 can be important to
  society

(Kraemer 24)
25
Easy Methods to Avoid Doing the Math Yourself

• http//www.surveysystem.com/sscalc.htm
• http//www.ezsurvey.com/samplesize.html
• http//www.macorr.com/ss_calculator.htm

26
References

• Frankfort-Nachmias, Chava and David Nachmias.
  Research Methods in the Social Sciences. 7th ed.
  New York Worth Publishers, 2008.
• Kraemer, Helena Chmura and Sue Thiemann. How Many
  Subjects? Statistical Power Analysis in
  Research. Newbury Park, Calif. Sage
  Publications, 1987.
• Powell, Ronald R. and Lynn Silipigni Connaway.
  Basic Research Methods for Librarians. Library
  and Information Science Text Series. 4th ed.
  Westport, Conn. Libraries Unlimited, 2004.
• Rosenthal, Robert and Ralph L. Rosnow. Essentials
  of Behavioral Research Methods and Data
  Analysis. 3rd ed. Boston McGraw-Hill, 2008.

27
Sample Error and Sampling Bias
Terence Kratz IS280 - 11/12/09
28
Sample Error

• How close the sample size is to the population
  size, or how well a sample of that size
  approximates a given population.

29
What is it?

• The standard deviation of a sampling distribution
  is referred to as the standard error or sampling
  error.
• Margin of Error
• The greater your sample size, the smaller the
  standard error. This is because the greater the
  sample size, the closer the sample is to the
  actual population itself.
• Varies depending on what is being sampled.

30
Why should you care?

• Depending on what claims are being made for what
  size population, the margin of error can indicate
  how strong the relationship being shown by the
  study actually is.

31
Sampling Bias

• The error resulting from taking a non-random
  sample of a population

32
What is it?

• Based on sampling method used, some members of a
  population are less likely to be included in the
  sample.
• May undermine the external validity of a test.
• Reduces the ability for results to be generalized
  to a larger population.
• Some studies might deliberately take a biased
  sample in order to produce misleading results.
• More often, sampling bias occurs because of the
  inherent difficulty in obtaining a truly
  representative sample of a complex population.

33
Types of Sampling Bias

• Selection from only a specific area of the
  population (intentional (purposive), or
  accidental convenience sample)?
• Self-selection bias
• Pre-screening of or advertising for volunteers
  within particular groups
• Exclusion bias

34
Selection from a Specific Area

• Biased if certain members are under-represented
  relative to others in the population being
  generalized to.
• Biased if certain members are overrepresented
  relative to others in the population.
• Appropriateness depends on the study and the
  population also called non-probability
  sampling.

35
Pre-screening

• Related to Selection from a Specific Area, but
  this form of bias occurs when participants are
  specifically screened for certain
  characteristics.
• Often used intentionally to focus a study on
  segments of the population with certain traits
  (ex. medical drug trials).
• May also include selecting only from certain
  kinds or groups of subjects in order to
  intentionally skew the sample toward a certain
  desired trait or characteristic.

36
Self-Selection Bias

• Possible whenever the group of people being
  studied has any form of control over whether to
  participate.
• Participants' decision to participate may be
  correlated with traits that affect the study,
  making the participants a non-representative
  sample.
• People who have strong opinions or substantial
  knowledge may be more willing to spend time
  answering a survey than those who do not.

37
Self Selection Bias Cont. Polls

• Online and phone-in polls, which are biased
  samples because the respondents are
  self-selected. This means that people with strong
  opinions are more likely to respond to the poll
  than people who have less strong opinions (or
  feel indifferent). This tends to polarize
  responses by giving greater weight to segments of
  the population with extreme opinions.

38
Exclusion bias

• Exclusion, often inadvertent, of any particular
  group or type of subjects from the sample.
• e.x. exclusion of subjects who have recently
  migrated into the study area (population
  demographics have changed since the study was
  initiated).
• Intentional exclusion is also sometimes used to
  screen out types of subjects that would normally
  be expected to be outliers in the study.

39
Literature Review

• Mixed Method Sampling
• Onwuegbuzie, Anthony J. and Kathleen M. T.
  Collins. A Typology of Mixed Methods Sampling
  Designs in Social Science Research," The
  Qualitative Report 12, no. 2 (2007) 281-316
• sam bloomberg-rissman

40
Mixed method sampling
QUALITATIVE AND QUANTITATIVE METHODS USED

• Onwuegbuzie, Anthony J. and Kathleen M. T.
  Collins. A Typology of Mixed Methods Sampling
  Designs in Social Science Research," The
  Qualitative Report 12, no. 2 (2007) 281-316

41
Generalizations

• Onwuegbuzie, Anthony J. and Kathleen M. T.
  Collins. A Typology of Mixed Methods Sampling
  Designs in Social Science Research," The
  Qualitative Report 12, no. 2 (2007) 281-316

42
Sample Myths

• Onwuegbuzie, Anthony J. and Kathleen M. T.
  Collins. A Typology of Mixed Methods Sampling
  Designs in Social Science Research," The
  Qualitative Report 12, no. 2 (2007) 281-316

43
Sample Method Selection

• Onwuegbuzie, Anthony J. and Kathleen M. T.
  Collins. A Typology of Mixed Methods Sampling
  Designs in Social Science Research," The
  Qualitative Report 12, no. 2 (2007) 281-316

44
Sample Size?

• Onwuegbuzie, Anthony J. and Kathleen M. T.
  Collins. A Typology of Mixed Methods Sampling
  Designs in Social Science Research," The
  Qualitative Report 12, no. 2 (2007) 281-316

45
Mixed Methods?

• Onwuegbuzie, Anthony J. and Kathleen M. T.
  Collins. A Typology of Mixed Methods Sampling
  Designs in Social Science Research," The
  Qualitative Report 12, no. 2 (2007) 281-316

46
Concurrent? Sequential?

• Onwuegbuzie, Anthony J. and Kathleen M. T.
  Collins. A Typology of Mixed Methods Sampling
  Designs in Social Science Research," The
  Qualitative Report 12, no. 2 (2007) 281-316

47
Representation and Legitimation

• Onwuegbuzie, Anthony J. and Kathleen M. T.
  Collins. A Typology of Mixed Methods Sampling
  Designs in Social Science Research," The
  Qualitative Report 12, no. 2 (2007) 281-316

48
Ethical sampling

• Onwuegbuzie, Anthony J. and Kathleen M. T.
  Collins. A Typology of Mixed Methods Sampling
  Designs in Social Science Research," The
  Qualitative Report 12, no. 2 (2007) 281-316

49
Conclusion

• Onwuegbuzie, Anthony J. and Kathleen M. T.
  Collins. A Typology of Mixed Methods Sampling
  Designs in Social Science Research," The
  Qualitative Report 12, no. 2 (2007) 281-316

50
REFERENCES

• Babbie, Earl. The Practice of Social Research.
  Belmont Wadsworth Cengage Learning, 2009.
• Frankfort-Nachmias, Chava and David Nachmias.
  Research Methods in the Social Sciences. 7th ed.
  New York Worth Publishers, 2008.
• Kerlinger, Fred N. Foundations of Behavioral
  Research. Fort Worth Harcourt Brace Jovanovich
  College Publishers, 1986.
• Kraemer, Helena Chmura and Sue Thiemann. How Many
  Subjects? Statistical Power Analysis in
  Research. Newbury Park, Calif. Sage
  Publications, 1987.
• Nachmias, David and Chava Nachmias. Research
  Methods in the Social Sciences. New York St.
  Martins Press, 1976.
• Onwuegbuzie, Anthony J. and Kathleen M. T.
  Collins. A Typology of Mixed Methods Sampling
  Designs in Social Science Research," The
  Qualitative Report 12, no. 2 (2007) 281-316
• Powell, Ronald R. and Lynn Silipigni Connaway.
  Basic Research Methods for Librarians. Library
  and Information Science Text Series. 4th ed.
  Westport, Conn. Libraries Unlimited, 2004.
• Rosenthal, Robert and Ralph L. Rosnow. Essentials
  of Behavioral Research Methods and Data
  Analysis. 3rd ed. Boston McGraw-Hill, 2008.
• Rosenthal, Robert and Ralph L. Rosnow. Essentials
  of Behavioral Research Methods and Data
  Analysis. New York McGraw Hill Inc., 1991.
• Trochim, William M. K. Sampling
  http//www.socialresearchmethods.net/kb/sampling.p
  hp Research Methods Knowledge Base (accessed
  November 11, 2009).
• Photos on slides 39-49 courtesy Sam
  Bloomberg-Rissman Photography