Mgt 540 Research Methods Sampling Issues
Basic Research Progress Explorative - Descriptive (Qualitative) Framework / Domain extant knowledge for reference Research Design Data collection / presentation Data analysis Emergent themes Relationship to extant knowledge? Tie to lit review, other research Findings? Possible hypothesis? Hypothesis Testing (Quantative) Framework / Domain Foundation Conceptual framework Hypothesis presentation Research Design Data collection / presentation Data analysis Confirm/disconfirm Findings? Possible additional hypotheses?
Research Design Flowchart FROM CHAPTER 11 Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
Sampling Why sample? Budget restrictions Time constraints Inaccessibility of some population members Sufficient accuracy, reliability with good sample Larger sample required for more heterogeneous population Randomly chosen sample is fair in the sense that every member of the population has an equal chance of being chosen
Sampling issues (terms) Selection of sufficient number of items or elements so that the properties of the sample (statistic) could be generalized to the population (parameter) Population Frame Listing of population elements Population Entire group of interest to researcher (people, things, events) Sample Subgroup of the population Subject Single member of a sample Element Single member of the population
Sampling precision Precision Degree of sampling error Measured by the standard error of the estimate See page 286 and Statistical tables, beginning on page 432
Sampling = Sample Mean S = Std. Deviation μ= Population Mean FIGURE 11.1 Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
Questions for determining sample Relevant target population? Exact parameters of interest? Kind of sampling frame available? Sample size needed (for desired level of confidence)? Cost relating to sampling design? Time available to collect data from sample? FIGURE 11.2 Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
Sampling Frame The empirical representation of the theoretical universe of interest In theory may be the entire population But, for example Not all own telephones (for a telephone survey) Some may be homeless (for a mail survey)
Sampling Unit Compare to the desired unit of analysis Individuals Dyads Work groups, teams Companies Industries Markets
Sampling Issues 11B Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
Probability & Non Probability Sampling 11C Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
Simple Random Sampling Most representative for most purposes Disadvantages Cumbersome and tedious Entire listing of all elements in the desired population are usually not available Very expensive Not the most efficient design
Complex probability sampling(s) Systematic Stratified random sampling Cluster sampling Area sampling Double sampling
Systematic sampling Every nth element is sampled, starting from a randomly chosen element 11F Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
Stratified random sampling Number of mutually exclusive sub-populations or strata e.g. university students divided into juniors, seniors, etc. Homogeneity within stratum and heterogeneity between strata Statistical efficiency greater in stratified samples Sub-groups can be analyzed Different methods of analysis can be used for different sub-groups 11G Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
Stratified random sample 11H Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
Stratified random sample TABLE 11.1 Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
11I Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
11J Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
Non-Probability Sampling Convenience samples the researcher’s convenience – unrestricted Purposive samples Judgment sampling – expert selection of respondents Quota sampling – ensuring representation of certain groups, individuals Snowball sampling – initially selected respondents (by probability or not) refer later ones
11K Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
11L Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
Precision 11M Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
Sampling considerations What is the relevant population? What type of sample should be drawn? What sampling frame should be used? What are the parameters of interest? How much accuracy and precision are desired? What is the sample size needed? What are the sampling costs? 11N Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
Sample size considerations Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
Sampling Efficiency Using n = sample size, S = standard error Efficiency is achieved when: Keeping n constant, you achieve a smaller S Reduce n keep the same level of S 11P Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
FIGURE 11.3 Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
Precision vs. Confidence More Precision Less Confidence More Confidence Less Precision FIGURE 11.4 Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
Pg. 294 TABLE 11.3 Copyright © 2003 John Wiley & Sons, Inc. Sekaran/RESEARCH 4E
Relevance of sample size Refer back to diagram on page 175 Purpose of Research? Exploratory Discovery Hypothesis testing? Types of investigation? Differences? Correlations? Causality? Unit of analysis? Data Collection method? Qualitative? Quantitative? Measurement / Measures?
Sampling exercises “What kinds of sampling designs for….” A study to get a quick idea of the medical acceptability of a new aspirin substitute which cannot be dispensed over the counter without prescription. Purposive judgment sampling
Sampling exercises “What kinds of sampling designs for….” A study involving a sample of 325 students in a university where 2,000 students are enrolled. A systematic sampling design (using a university listing of students An investigation of the career salience of professionals in the fields of medicine, engineering, business, and law. A stratified random sampling with stratification along profession, gender, age, etc.
Sampling exercises “What kinds of sampling designs for….” The generalizability of the attitudes of blue collar workers from a sample of 184, to the total population of 350 blue collar workers in the entire factory of a particular company. Simple random sampling (because of the high importance attached to generalizability
Sampling exercise (problem) You want to estimate the production days that would be lost during the next three months by sampling the vacation intentions of a few employees. You randomly select 36 employees in the organization and find that the average number of days they intend taking off is 16 during the coming three Summer months, with a standard deviation of seven (7) days. Based on these sample statistics, you want to estimate at a 99 percent confidence level, the days that will be lost due to the entire population of workers taking vacation time during the next three months, so that the plant manager knows how much temporary help he should plan on hiring during the summer months in order for work to proceed smoothly.
Exercise calculation Solution: μ = ± z S S = S/√n = 7/6 = 1.167 = 16 ± 3.01 = 12.99 to 19.01 = Sample Mean μ = Population Mean S = Std. Deviation S = Standard Error
If there are 100 employees in the organization expected to take vacation during Summer, then, the most optimistic estimation of the days lost through vacation time during the summer would be (13 x 100 =) 1m300 days and the most pessimistic would be (19 x 100 =) 1900 days. This would mean that temporary help would be needed anywhere between 1,300 and 1,900 days worth of labor for production to proceed smoothly. To narrow the gap – (increase precision) requires sacrificing confidence – choose your risk.