2-5 STUDY POPULATION AND SAMPLING Subgroup or portion of the population
Represents the population which is based SAMPLE SURVEY any statement about the population from simply referred as “survey” which it is drawn just a part of the population who is CENSUS answering the questionnaire Count or study of every unit or member of a Sample –it is considered to be a part of the given population population which is representative of the Determination of the distribution of their characteristics of the population. characteristics, information obtained for Survey –economical and practical than the each of the units census SAMPLING DESIGN o Census –is a population study. Scheme, strategy or procedure of arriving at Everybody will be included as the sample which involves specification of respondents. the target population, the respondent ADVANTAGES OF SURVEY population and the method of selecting the Faster and cheaper sample More comprehensive information MARGIN OF ERROR –refers to the limit of More accurate accuracy Much larger and much variated 1% = 0.1 (99% accurate) 5% = 0.05 (95% accurate) SAMPLING 10% = 1 (90% accurate) Process of selecting individuals from the PROBABILITY total population to be studied Refers to the level of confidence in Ensures that the sample is representative of investigator has on the accuracy of his the study population finding which may be set statistically at Representativeness (main goal) certain percentage 5% margin of error-probability of 95% POPULATION SAMPLE SIZE All members, units or elements of any well- Number of samples to represent the defined set of group of people or objects population under study which should be which is the focus of the investigation and adequate to warrant generalization of the from which a sample is drawn. findings CRITERIA IN SAMPLING STRATUM Defined by one or more specification that ADEQUACY OF SAMPLE divides a population into mutually exclusive Quantity or number of required sample to components or parts represent the population Computation using sample size formula TARGET POPULATION Group or set of individuals, check, or events REPRESENTATIVENESS OF THE which is the subject or focus of the SAMPLE investigations. Sample mix characteristics which are similar or identical to those of the population from HOMOGENOUS where it was taken If the units or members have identical or A miniature of the population – just like it, similar characteristics like age, sex, marital only smaller status, economic status, religion, education and ethnicity. RANDOMIZATION HETEROGENOUS Process of assigning subjects to control If the units or members have extremely groups or to groups which are to be diverse or different characteristics. compared in a non-experimental
SAMPLE MARINILL P. SOLIMAN 3-H
2-6
DETERMINING THE SAMPLE SIZE Where:
o K is the sampling interval Desired confidence (alpha) o N is the total number of the o 0.01 (1%) population o 0.05 (5%) o n is the total number of the o 0.10 (10%) sample 3. STRATIFIED SAMPLING 2 Heterogeneous population N Z −p (1− p) n= Ensures adequate representations for each of N e 2 +Z 2 p (1− p) the population subgroups or strata in the sample Where Z is the confidence level value, as follows: Stratified simple random sampling 99% confidence level Z = 2.58 Stratified systematic sampling Stratified proportion sampling 95% confidence level Z = 1.96 4. CLUSTER SAMPLING 90% confidence level Z = 1.65 Random selection of groups, not individuals, p is the largest possible proportion equals 0.50 as samples (assumed preliminary estimate) Cluster refers to any entire group of n = sample size individuals or objects which share similar N = population size characteristics important in a study e = sampling error Area sampling Geographical consideration SAMPLING DESIGNS 5. MULTI-STAGE SAMPLING Several stages or phases Slovin’s formula: Population units are grouped and arranged N into hierarchal order or level n= 1+ N e 2 Sampling is done successively sample regions and provinces 2 MAJOR TYPES ADVANTAGES Less bias Probability sampling –random-sampling Equal chance to be selected technique. DISADVANTAGES Equal probability that it will be included in Time consuming the sample Expensive Fair sample selection and allows estimation Inconvenient Non-probability sampling –non-random NON-PROBABILITY SAMPLING DESIGNS sampling technique No way of estimating the probability that 1. Purposive sampling each unit has a chance of being included in Good judgment and appropriate strategy the sample 2. Convenience sampling Convenience and economical “whoever is available” constitute the sample and continues doing so until the PROBABILITY SAMPLING DESIGNS predetermined sample size is reached 3. Quota sampling 1. Simple random sampling Equivalent of stratified sampling under Basic, best known probability sampling the probability category design Selection of the subjects to constitute Equal chance of being included in the the sample in each stratum is done in a sample non-random manner 2. Systematic sampling 4. Snowball sampling No available list of population units to Establish the sampling interval work on Requires identification of a few persons N who meet the requisite characteristics K= important to the study n MARINILL P. SOLIMAN 3-H
Researchers must be cautious in the sample
ADVANTAGES selection process in order to eliminate or Convenient control extraneous variables Easy to produce anytime SAMPLING AND STATISTICAL ANALYSIS Less time spent for selection of the Inferential statistics in sample size respondents Statistics in complete enumeration Economical Less expensive No fixed budget required DISADVANTAGES Produce biased samples or errors in judgment No estimation of precise elements of the population that will be included in the samples Not equal chances 5. Theoretical sampling An approach used in grounded theory studies Aims to advance the development of a selected theory Researcher’s goal of theoretical sampling is not the same as with probability sampling nor the representative Can be viewed as a technique of data triangulation using independent pieces of information UNIVERSE Is a totality of elements to which research findings may apply Target population People or objects from which the researcher intends to collect data and generalize the findings of the study ELEMENTS The entities that make up the sample and the population SAMPLING BIAS Is influenced by external factors so that the data no longer represents the entire population STEPS IN SAMPLING 1. Identify the target population or the universe 2. Identify the respondent population 3. Specify the inclusion and exclusion criteria for respondent selection 4. Specify the sampling design 5. Recruit the subjects INCLUSION CRITERIA Guidelines choose fitting subjects with a predetermined set of characteristics Demographic, geographic and temporal criteria as appropriate INCLUSION CRITERIA Characteristics that exclude potential subjects from the study Not suitable for the study even though they meet the inclusion criteria