2) Sampling

Sample Size No simple rules...... 1. Rule of 10 . . . 2. Rule of 30 . . .

1. 10 elements (participants) per each variable 2. Each group should have 30 elements (participants)

1. Simple random sampling to select groups/ "clusters" and then subjects selected from within clusters o "clusters" become the unit of analysis 2. What key words might you find?

1. Cluster Sampling "Multistage Sampling" 2. Universities, Hospitals, Nursing Homes

1. recruiting all member of a population over specified period of time o Aka Rolling enrollment 2. What type of design is this?

1. Consecutive sampling 2. Non-probability sampling

what type of sampling method?? 1. The clients in the hypertension clinic of a local hospital are studied.____________ 2. A total of 20 nursing service administrators are randomly selected from a random sample of 10 hospitals in the state_______ 3. Every 5th nurse is randomly selected from the mailing list of the ANA_____________

1. Convenience 2. Cluster 3. Systematic Random Sampling

True or False 1. Attrition has little effect on generalizability of study findings 2. A study involving 2 variables needs a minimum of 14 subjects 3. A sample with elements having similar attributes would be considered homogenous 4. The use of power analysis is the best way to determine sample size in a quantitative study

1. F 2. F 3. T 4. T

1. ______ : degree to which elements are similar o Ex: Stage 4 renal failure on dialysis 2. ______: degree to which they are diverse o More variability, larger sample size you will need.

1. Homogeneity 2. Heterogeneity

1. What three things do you need for probability sampling? 2. What is the difference between random sampling and random assignment?

1. Identifiable accessible population, sampling frame, and random selection from sampling frame. 2. Random sampling is a way to get people recruited. Random assignment is research design. Randomizing the people who have already consented to experimental or control group

1. Elements must possess characteristics to be included in sample 2. Element possesses characteristics not desired in study 3. Why is #2 important?

1. Inclusion criteria 2. Exclusion criteria 3. Can exclude people who could be included but have confounding factors

1. Which designs are less lilley to produce representative samples? 2. Give an example of this type of design

1. Non-probability samples 2. Convenience samples

1. _____ sampling - only method to obtain representative sample best way 2. _____ sampling -offers ease and economy, often more practical not strongest

1. Probability 2. Nonprobability

1. Researchers hand pick sample members. Participants are particularly knowledgeable about study subject. 2. What type of design is this? 3. Is this usually for qualitative or quantitative studies?

1. Purposive sampling 2. Non-probability sampling 3. Qualitative studies

what type of sampling method? 1. The first 30 men and the first 30 women who are admitted to the hospital for orthopedic surgery during the time of the research study are asked to participate____________________ 2. To determine the frequency of patient falls, a sample of 100 charts is randomly selected from all of the patient charts during the previous six months.__________________

1. Quota 2. Simple Random Sampling

1. Identify the strata (groups in the sample) and determine how many are needed per strata. Enroll until quota is filled (can be form of convenience sampling) 2. What type of design is this?

1. Quota sampling 2. Non-probability sampling

1. selecting a portion of the accessible population to represent the population 2. a sample with characteristics that closely resemble the greater population 3. segments of the population that have characteristics that are mutually exclusive

1. Sampling 2. Representative sample 3. Strata

1. ____ - "chance" or "standard error" o Elements do not sufficiently represent population o Caused by chance variation in population o Not under researcher's control Can even happen w/ good methods 2. ______ - Systematic over or underrepresentation of a segment of the population relating to particular study characteristic(s). o Researcher control o Threat to External Validity/Generalizability

1. Sampling Error 2. Sampling Bias

Components of Power Analysis 1. What are the 4? 2. _____ - Traditionally set at 0.05 -5/100 due to chance (error) Talk about next week. Risk of making error - 5% 3. _____ - Moderate selected unless researcher has reason to believe otherwise 4. _____ -Acceptable norm is 0.80 (or higher) 5. _____ - As size increases, power increases

1. Significance level (alpha), Effect size, Power, and Sample size. SEPS 2. Significance level (alpha) 3. Effect size 4. Power 5. Sample size

1. Previously identified members of a group identify other members 2. What type of design is this? 3. Is this usually for qualitative or quantitative studies?

1. Snowball/Networking 2. Non-probability sampling 3. Qualitative studies

1. _____ - "the capacity to detect true relationships" 2. How is this achieved? 3. How do you do this? 4. Expected difference is LARGE > sample can be larger/smaller 5. Expected difference is small> sample should be larger/smaller 6. What if you don't know the effect size?

1. Statistical power 2. Adequate sample size 3. Power analysis 4. smaller 5. larger 7. Expect it to be a small difference and get a large sample

True or False 1. A sampling frame is a listing of all elements of a population. 2. Non-probability sampling means there is no probability that the subjects selected will constitute a biased sample. 3. The best means of obtaining an unbiased sample of subjects in a community is to select a random sample of names from the telephone directory.

1. T 2. F 3. F

1. ____ - statistically significant difference between groups when there really ISN''T one o Erroneous rejection of the Null Hypothesis 2. ____ - no statistically significant difference between groups when there really IS one o Erroneous acceptance of Null hypothesis

1. Type I Error 2. Type II Error

Determination of the appropriate sample size in quantitative study is based on the principle(s) of: A. Power analysis B. Saturation and redundancy C. Convenience D. The "rule of 10"

A

Sampling criteria may be defined as A. The identification of eligibility or ineligibility to participate in a study B. The population in which the investigator is interested C. The selection of a subset of a population to represent the whole population D. A technique used to ensure that each person has a chance of being included in a study

A

What are these describing? Advantages - Used when it is difficult to determine sample frame in advance - Less time and money disadvantages - Boas from use of convenience sampling after stratification - May not be truly representative of the target pop.

Advantages and Disadvantages of Quota sampling

What are these describing? Advantages - Ease of data collection - Economical and time saving Disadvantages - Weakest method of sampling - Highest probability of selection bias

Advantages and Disadvantages of a Convenience sample

A total of 20 nursing students are randomly selected from a random sample of 5 nursing programs in one state. This is an example of: A. Simple random sampling B. Cluster sampling C. Convenience sampling D. Purposive sampling

B

In interpreting quantitative research results, the representativeness of the sample is most closely tied to A. Internal Validity B. External Validity C. Sample Validity D. Research Validity

B

Power analysis is conducted to : A. Determine a large effect size B. Estimate sample size C. Test for internal validity D. Set the level of significance

B

Which of the following samples is least likely to be representative of the overall population? A. Convenience B. Consecutive C. Simple Random D. Stratified Random

B

Which of the following studies would require the largest sample size A. A study with a homogenous study sample B. A study with an alpha of 0.01 C. A study with a large effect size D. A study with few variables

B

A study sampling frame is 250. Power analysis suggests a sample size of 50. What is the sampling interval?? A. 25 B. 50 C. 5 D. 10

C

A researcher has decided to conduct a satisfaction survey among all of the patients who presented ambulatory surgery unit over a two-month period of time. This is an example of: A. Stratified random sampling B. Cluster sampling C. Convenience sampling D. Purposive sampling

C (consecutive is a type of convenience sampling)

What is this describing? Advantages - Economical, time saving w/ large dispersed population - Characteristics of clusters as well as population can be estimated Disadvantages - Likelihood of sampling error increases with each sampling stage - Statistical analysis complex - Tends to be less accurate than simple or stratified

Cluster Sampling

Another name for cluster sampling is: A. Systematic B. Quota C. Network D. Multistage

D

What is this describing? Population is divided into strata and then elements are randomly selected from strata

Stratified random sampling

What is this describing? Advantages - Saves time and money - Easy to draw sample Disadvantages - If population ordering is not random, it can introduce bias - After 1st element selected, others no longer have equal chance of selection

Systematic Random Sampling

What is this describing? Every 5th (or interval of choosing) from a list, or frame Can be random Establish sampling interval (population 1,000, need 100 sample. Every 10th case selected. Choose random #-8. Select case 8, 18, 28, etc)

Systematic sampling

Where do you find the sample in a journal?

Under methods