MSU Sports Interests ObjectivesPopulation/SampleVariables/Info neededData gatheringSlide 5Slide 6Sampling error formulaSampling errors for binomial (95% confidence interval) percent distribution in populationSample problemsMSU Sports InterestsObjectives•Measure student participation in various sports on and off -campus•Measure student interests in participation for different activities & levels of competition•Compare partic & interests of men and womenPopulation/Sample•All MSU students enrolled spring semester 1978•Stratify sample by class standing•seek sample of 200 men, 200 women–assume 67% response rate–mail out 600 surveysVariables/Info needed•Gender, age, class •Participation in sports–ever, last year, on campus– 40 activities•Interests in sports partic on campus–40 activities–levels of competitionData gathering•Self-administered•mailed 4 page questionnaire•no follow-ups1. Define problem and study objectives2. Identify information needs & study population(s)3. Determine basic design/approach- cross sectional vs longitudinal- on-site vs household vs other- self-admin. vs personal interview vs phone- structured or unstructured questions4. Questionnaire design 5. Choose sample (frame, size, sampling design) 6. Estimate time, costs, manpower needs, etc. STEPS IN A SURVEY7. Proposal & “Human subjects” review 8. Line up necessary resources 9. Pre-test instruments and field procedures 10. Data gathering and follow-up procedures11. Coding, cleaning and data processing 12. Analysis: preliminary, then final. 13. Communication and presentation of results. Survey ImplementationSampling error formulan = Z2 2 / e 21. Solve for e to express error as a function of sample size, confidence level, and variance: e = (Z * ) / SQRT ( n )2. For binomial, = sqrt (p(1-p)) , where p is proportion for “yes” in the populationGenerate numbers in binomial sampling error table as:[1.96 *sqrt( p * (1-p)) ]/ sqrt (n)Sampling errors for binomial (95% confidence interval)percent distribution in populationSamplesize50/50 60/40 70/30 80/20 90/10100 10.0% 9.8% 9.2% 8.0% 6.0%200 7.1% 6.9% 6.5% 5.7% 4.2%400 5.0% 4.9% 4.6% 4.0% 3.0%1000 3.2% 3.1% 2.9% 2.5% 1.9%1500 2.6% 2.5% 2.4% 2.1% 1.5%2000 2.2% 2.2% 2.0% 1.8% 1.3%Sample problems•From binomial table: Survey with sample of 400 estimates 32% of sample have visited park X. Find 95% confidence interval for population estimate.•Using formula: Same survey (n=400) estimates average number of visits per year to be 5.6. If standard deviation in sample is 4, find 95% CI for this
View Full Document