PR II Sampling Methods PDF

SAMPLING METHODS & TECHNIQUES PR II | Karen Collins Ramos LEARNING GOALS 1 2 3 Discuss relative Distinguish between Differentiate sampling adva...

SAMPLING METHODS & TECHNIQUES PR II | Karen Collins Ramos LEARNING GOALS 1 2 3 Discuss relative Distinguish between Differentiate sampling advantages and probability & methods and techniques disadvantages of each non-probability sampling sampling method  01 Introduction to Sampling  RESEARCH GENERALIZABILITY TERMINOLOGIES 01 02 Target Population Study Population To whom do you want to When the target population is hard to access, you generalize the results? can get what is available from the target population 03 04 Sampling Frame Sample List of all elements of the study Actual selection for your population that will be used for study sampling TERMINOLOGY SUMMARY TARGET POPULATION To whom do you ideally want to generalize the results? STUDY POPULATION To what population can you realistically gain access? SAMPLING FRAME How can you gain access to them? SAMPLE Who is in your study?  EXAMPLE 01 01 02 TARGET POP. All elderly people STUDY POP. with Alzheimer’s All elderly people with 03 Alzheimer’s in QC 04 hospitals SAMPLING FRAME A list of all elderly people with SAMPLE Alzheimer’s in QC homes Actual people with Alzheimer's who participated in your study  EXAMPLE 02 01 02 TARGET POP. School-age children with STUDY POP. asthma School-age children with 03 asthma treated in 04 pediatric centers in Metro Manila SAMPLING FRAME A list of school-age children with SAMPLE asthma treated in pediatric centers Actual children with in Metro Manila asthma in Metro Manila who participate in your study  02 Sampling Basics SAMPLE Smaller (but hopefully representative) collection of units from a population used to determine truths about that population (Field, 2005) Consists of elements from the study population that are actually selected to participate WHY SAMPLE? LIMITED RESOURCES CALCULATE ACCURACY Time Gives results with known accuracy that Money can be calculated mathematically Workload POSSIBLE POPULATIONS All female All doctors All Gen Z teens All Filipino high School children Men aged 15-21 school students SAMPLING POPULATION SAMPLE SAMPLING POPULATION SAMPLE FACTORS THAT INFLUENCE SAMPLE REPRESENTATIVENESS Sampling Procedure Sample Size Participation (response)  SAMPLING RULE OF THUMB: the larger the sample size, the better  SAMPLING Question: What happens mathematically if there are too many participants or data points? WHEN MIGHT YOU SAMPLE THE ENTIRE POPULATION? When your population is very small When you have extensive resources When you don’t expect a very high response  03 CONSIDERATIONS in SAMPLING How do you determine the number of participants in your study? O1 Consideration: CONFIDENCE LEVEL Confidence Level (90%, 95%, 98%, 99%) ○ Level of certainty that your data is representative of the entire population Ex. 95% confidence level = 95% certain that the results reflect the opinions of the entire population 02 Consideration: MARGIN OF ERROR ○ Also known as confidence interval ○ Shows what the uncertainty is ○ Tells you how much you can expect your results to reflect the views from the overall population ○ Ex 4% margin of error = your results are within 4 percentage points of the real population value  Confidence level is inversely proportional to the margin of error. ○ The larger the margin of error, the lower the precision rate. ○ The smaller the margin of error, the higher the precision rate. BIG IDEAS  Larger sample sizes Effect Size Closer to the Small Sample Risk The more data you have, the more population size Possibility of the sample capacity to draw out existing being unusual by chance relationships between them Less of an inference if the sample size is large  04 Sampling Bias & Sampling Error  Getting responses only from those who are interested or available SAMPLING BIAS Over-representation & underrepresentation EXAMPLE Study habits of Grade 11 students in Metro Manila Readiness and attitudes towards online learning  Results when there is sampling bias ○ Samples are not representative of the target population SAMPLING ERROR ○ Normally corrected by increasing the sample size NON SAMPLING ERROR Problems in data collection or processing ○ Low response rate ○ Error in instrument in data collection (validity and reliability) ○ Mistakes in data encoding  05 Sampling Techniques PROBABILITY Measure of the likelihood that an event will occur in a random experiment Quantified as a number between 0 and 1, where 0 = impossibility and 1 = certainty Meaning: the higher the probability of an event, the more likely it is that the event will occur TYPES OF SAMPLING PROBABILITY NON-PROBABILITY Random selection Judgment of researcher Every member of a population has a The odds of any member being selected for known (calculated) and equal chance of a sample cannot be calculated being selected *does not mean odds are equal [ex. 1 person has 10% chance of being selected and another has 50% chance] WHICH WOULD YIELD A REPRESENTATIVE SAMPLE? PROBABILITY NON-PROBABILITY Population: 100 Population: 100 Probability: Each person will have Probability: Cannot be calculated odds of 1 out of 100 of being chosen  PROBABILITY OR NON-PROBABILITY? COLLEGE ADMISSIONS  Any method that uses some form of random selection Process or procedure that assures that different units in your population have equal probabilities of being chosen Traditional: picking names from a hat PROBABILITY [now, computer-generated] SAMPLING PROBABILITY SAMPLING 01 02 SIMPLE RANDOM STRATIFIED RANDOM 03 04 SAMPLING SYSTEMATIC RANDOM CLUSTER SAMPLING RANDOM SAMPLING 01 RANDOM SAMPLING Purest form of probability sampling Each member of a population has an equal and known chance of being selected ○ PROCESS: randomly assign a number to all participants and use a random number generator to choose random numbers Case of large populations [sampling frame] 02 STRATIFIED RANDOM SAMPLING Stratum: subset of population that shares at least one common characteristic ○ Ex. males/females, managers/non-managers Split members of a population into mutually exclusive groups then use simple random sampling to choose members from the groups  02 STRATIFIED RANDOM SAMPLING PROCESS ○ Identify relevant stratums and action representation in the population ○ Use random sampling to select sufficient number of subjects in each stratum Sufficient sample size: large enough for us to be reasonably confident that the stratum represents the population  02 STRATIFIED RANDOM SAMPLING EXAMPLE ○ Performance of high school students in the 2021 Ateneo College Admissions Process ○ Divide into two strata Public school students Private school students Assume sample size is 400, then ADMU must take 200 from each stratum  02 STRATIFIED RANDOM SAMPLING EXAMPLE ○ SC needs to survey 100 students Get random samples of 25 students in each batch 03 SYSTEMATIC RANDOM SAMPLING Nth name selection technique Means you choose every “nth” participant from a complete list. ○ Ex. choose every 10th person listed 03 SYSTEMATIC RANDOM SAMPLING PROCESS ○ Access list of all members of a population ○ List must not have any hidden order (random) ○ Assign n value (random number) that becomes nth element ○ Select every nth in the list for your sample 03 SYSTEMATIC RANDOM SAMPLING EXAMPLE ○ Population: MIIS Upper School students ○ Get list of names from the Registrar Make sure list is not arranged in any way (ex. Ranking, alphabetical order, intelligence etc) ○ Select every 5th name on the list  04 CLUSTER RANDOM SAMPLING Obtained by dividing the study population into clusters (typically geographically) Ideally, members of the clusters are homogenous/similar A cluster sample gets every member from some of the groups Used when it is not possible to get a list of all units of a population (ex. All Filipinos) 04 CLUSTER RANDOM SAMPLING PROCESS ○ Get a list of all clusters of the population ○ Randomly select a number of clusters as your sample Take note that all members in the cluster will be in your sample 04 CLUSTER RANDOM SAMPLING EXAMPLE ○ Buying capacity of Filipinos for marketing of a product Randomly select a certain number of regions (clusters) as target population Survey all participants in the selected clusters WHAT ARE KEYWORDS TO REMEMBER IN EACH TYPE? 01 02 SIMPLE RANDOM STRATIFIED RANDOM 03 04 SAMPLING SYSTEMATIC RANDOM CLUSTER SAMPLING RANDOM SAMPLING  03 NON-PROBABILITY SAMPLING  Does not involve random selection or measures Odds of selecting a unit in the population cannot be calculated or is unknown NON-PROBABILITY SAMPLING NON-PROBABILITY SAMPLING 01 02 CONVENIENCE SAMPLING QUOTA SAMPLING 03 04 PURPOSIVE SAMPLING SNOWBALL SAMPLING 01 CONVENIENCE SAMPLING Used in exploratory research where the researcher is interested in getting an inexpensive approximation of the truth Sample is selected because it is convenient Often used in preliminary research efforts to get a gross estimate of the results 02 QUOTA SAMPLING Non-probability equivalent of stratified sampling Groups in the sample are proportional to the groups in the population Identify stratums and their proportions as they are represented in the population ○ Ex proportion of men and women in the population, then use that as basis for selection for your sample 03 PURPOSIVE SAMPLING The researcher chooses a sample based on their knowledge about the population and the study itself. Usually an extension of convenience sampling Participants are chosen based on the study’s purpose. 04 SNOWBALL SAMPLING Special non-probability method used when the desired sample characteristic is rare Relies on referrals from initial subjects to generate additional subjects Often used in qualitative research  04 CONCEPT REVIEW  QUESTION 1 Sampling bias causes sampling errors. A. True B. False QUESTION 2 Probability sampling is based on the idea that each member or unit of a population has an equal and known chance of being selected. A. True B. False QUESTION 3 Non-probability sampling is often based on judgment of the researcher, and may also be for convenience purposes. A. True B. False QUESTION 4 You are given a random list of movie patrons of a movie house in Katipunan, Quezon City. You decide to survey every fifth person on the ticket list and ask what genres of movie they watch every week. This is an example of what sampling? A.Random B.Stratified C.Cluster D.Systematic QUESTION 5 In estimation of immunization coverage in a province, data on seven children aged 12-23 months in 30 clusters are used to determine proportion of fully immunized children in the province. Give reasons why cluster sampling is used in this survey. QUESTION 6 A group of researchers would like to see if women aged 20-50 who underwent laser treatments in beauty clinics possessed narcissistic tendencies and high self-presentation. What sampling method would you recommend and why? QUESTION 7 You would like to determine the proportion of underpaid migrant workers in East London. Which sampling method would be best to use and why? QUESTION 8 A researcher wants to test for the differences in coping mechanisms between AIDS patients and Stage 4 cancer patients. Which sampling method must the researcher use? QUESTION 9 Suppose you are interested to see if Filipinos who regularly watch Kdramas and follow Kpop boy bands are significantly different from those who follow Western pop culture in terms of attitudes towards romantic relationships. Which sampling would be advisable to use and why? QUESTION 10 ___________________ occurs when every member of a population has an equal chance of being selected for a sample. a. stratified random sampling b. simple random sampling c. area probability sampling d. systematic sampling QUESTION 11 Stratified random sampling is the preferred strategy when: a. no sampling frame is available. b. only a part of the population is accessible to researchers. c. you want to include specific subgroups in the study. d. the population is very small. QUESTION 12 Systematic sampling can produce a very biased sample when: a. there is a structure to the sampling frame. b. the population is too heterogeneous. c. the population is too small. d. there is no available sampling frame. QUESTION 13 Convenience samples are frequently used in student research because they: a. are preferred by instructors. b. yield representative samples. c. are more appropriate for statistical analysis. d. take less time and money. QUESTION 14 Farmer Joe separates his apple tree farm into 10 regions. He counts the number of apples produced in just one of the regions and uses that estimate to predict the number of apples produced on the whole farm. This is _______ sampling. a. simple b. stratified c. cluster d. systematic QUESTION 15 Farmer Joe's apple tree farm is set up in 100 rows. He counts the number of apples produced on every 10th row to estimate the number apples produced on the whole farm. This is _____________ sampling. a. simple b. stratified c. cluster d. systematic