Chapter 10 Flashcards
Explain the difference between a sample and a census, and why a sample is preferred above a census:
Sample - A subset of a population (or universe).
Census – A survey whereby data is obtained from or about every member of the population of interest.
Sample survey preferred over a census survey because:
*More accurate – better quality data
*Takes less time
*Costs less
*More practical
Give the six steps in the sampling process:
- Define the population.
- Identify the sampling frame.
- Select the sampling methods.
- Determine the sample size.
- Select the sample elements.
- Gather data from designated elements.
Name and explain step 1 of the sampling process:
- Defining the population:
Population – total group of people from whom information is needed. - Sample unit = unit available to be selected for the sample.
- Sample elements = unit from which information is required (usually individuals). With whom do I want to speak?
- Extent = geographical boundaries of population
- Time = during which period/time?
Name and explain step 2 of the sampling process:
- Identifying the Sample Frame:
Sample frame - list of all the sample units available for selection at a stage of the
sampling process.
Sample is drawn from within the sample frame (list, index or any population
record)
Sample Frame shortcomings:
*Missing sample units
*Duplicate entries
*Foreign elements
Name and explain step 3 of the sampling process:
- Selecting the sampling methods:
Non-probability sampling methods:
- Convenience sampling
- Judgement sampling
- Snowball sampling
- Quota sampling
probability sampling methods:
- Simple random sampling
- Cluster sampling
- Stratified sampling
- Systematic sampling
Name and explain step 4 of the sampling process:
- Determining the Sample Size:
Sample error
Sample size
Methods of determining sample size
Statistical calculation of sample size
Name and explain step 5 of the sampling process:
- Selecting the Sample Elements:
-Selecting respondents.
-Develop clear guidelines & procedures for selection of respondents.
-Based on sampling method.
Name and explain step 6 of the sampling process:
- Gathering Data from Designated Elements
Name 4 types of sampling frames:
- Computerised registers
- Address directories, buyer’s guides, yearbooks.
- Membership lists
- telephone directories.
Give the needed requirements for a reliable sample frame:
- All the elements and strata of the population are represented.
- It must be up to date.
- The details of each entry must be complete and correct.
- There must be no duplication of entries.
Explain Convenience sampling:
*Sample drawn from readily accessible/available section of population
*Same place & time as researcher
*Not representative of population = not generalisable
*Cost-effective
*Researcher has little control over sample contents
*Distinct degree of bias
*Exploratory research – ideas/insight more important than scientific objectivity.
Explain judgement sampling:
Also known as purposive sampling.
*Sample elements selected subjectively & deliberately to be representative of population.
*Subject to researcher’s judgement.
*Not generalisable.
*Useful if large samples not necessary.
*Pre-testing questionnaires, pilot studies, exploratory studies.
Explain Snowball Sampling:
*Method of judgement sampling.
*Samples of special population needed.
*Respondents with specific characteristics deliberately selected.
*Respondents identify other similar individuals.
*Helps find difficult-to-identify respondents.
*High sample bias
*Reduces search cost
*Qualitative studies
Explain Quota Sampling:
*Convenience & Judgement sampling
*Classify population according to relevant characteristics (e.g. age, income etc.)
*Sample elements according to quota (proportion) in population
*No sample frame necessary
*Faster & cheaper for surveys vs probability sampling
*Less control variables = lower cost, greater risk of selection bias
*Interviewer judgement & fieldwork NB
*Bias & non-representative
Explain Simple Random Sampling:
*Units of population selected individually & directly via random process.
*Each unit has known & equal chance of selection.
1. Without replacements: take elements out one by one and don’t put back.
2. With replacement: element noted & put back before next element selected.
*Requires complete, accessible sample frames & sufficient information on elements.
*Table or random draw
Explain Systematic Sampling:
*Sample elements drawn systematically from complete list of population.
*1st student chosen via simple random sampling.
*Thereafter, calculate interval length (skip interval) and select students at intervals to
obtain desired sample size.
*Simple, flexible, cost-effective
*Lists should not have patterns/trends causing bias
Explain Stratified Sampling:
Strata = levels/layers
Step 1: heterogeneous population grouped into 2/more homogeneous strata that are
mutually exclusive & comprehensive
Step 2: Random sample drawn independently from each stratum via random or systematic sampling.
*More representative sample
*Requires knowledge of population composition (e.g. sex, age, income etc.)
Explain Cluster Sampling:
*When sample must be drawn from population with difficult/impractical/impossible to
compile sampling frame.
*Randomly selects elements of population in clusters (not individually)
Step 1: Cluster = Total population divided into mutually exclusive & comprehensive
groups.
Step 2: Sample = random sample of elements drawn from each cluster via 1)single phase or 2) two-stage sampling
*Requires knowledge of population’s composition
Define and differentiate between Sampling errors and Systematic (non-sampling) errors:
- Sampling errors
Difference between population value & sample value.
Error due to only surveying small portion of population. - Systematic (non-sampling) errors
-Observation/measurement errors
-Administrative errors or bias
-Reduce via thorough planning
What are random sampling errors?
Error that arises because we only survey a
small portion (sample) of the population
Can be reduced by increasing the sample
size, but cannot be eliminated unless the
entire population is surveyed.
Give four reasons for systematic errors:
Reasons for systematic errors:
-interviewer’s lack of conception (insight) and logic (reasoning skills);
-arithmetical errors
-misinterpretation of results and statistics;
-incorrect tabulation, coding, and reporting.
Which elements determine the maximum sample size?
Maximum sample size determined by:
*Study’s purpose & required precision
*Population size
*Available resources
*Report requirements & importance
Name and explain the two methods of determining sample size:
- Blind guessing:
- The researcher uses judgement and intuition to determine the sample size. - Statistical method:
- Uses statistical formulae to determine the sample size, which is based on these three criteria: - The required level of confidence
- The required precision (desired degree of accuracy of sample results)
- The standard deviation of the population.
