Research Design and Sampling Techniques PDF

UNDERSTANDING DATA & WAYS TO SYSTEMATICALLY COLLECT DATA PRACTICAL RESEARCH 2 | WEEK 2 | SECOND QUARTER | QUANTITATIVE RESEARCH DESIGN AND SAMPLING  0bectives: Chooses appropriate quantitative research design (CS_RS12-IIa-c-1) Describes sampling procedure and the sample (CS_RS12-IIa-c-2) Di...

UNDERSTANDING DATA & WAYS TO SYSTEMATICALLY COLLECT DATA PRACTICAL RESEARCH 2 | WEEK 2 | SECOND QUARTER | QUANTITATIVE RESEARCH DESIGN AND SAMPLING  0bectives: Chooses appropriate quantitative research design (CS_RS12-IIa-c-1) Describes sampling procedure and the sample (CS_RS12-IIa-c-2) Discussion about Research Design and Sampling  WHAT IS THE IMPORTANCE OF PROPERLY WRITING THE CHAPTER 1 AND CHAPTER 2 OF YOUR RESEARCH PAPER? HOW ARE THESE TWO CONNECTED? WHAT DO THESE PHOTOS REPRESENT? PURPOSE OF THE METHODOLOGY ► Describe the research design (or defend) ► Provideenough details so that a competent worker can repeat the experiments  1. Define Research Research Problem Process 7. Interpret And Report 2. Review The Literature 6. Analyze And 3. Formulate Present Data Hypothesis 5. Collect 4. Design Data Research RESEARCH DESIGN ► The overall strategy that you choose to integrate the different components of the study in a coherent and logical way. ► It constitutes the blueprint for the collection, measurement, and analysis of data. TERMS TO Replication KNOW Local Control Randomization Control of Extraneous Variables  QUANTITATIVE RESEARCH DESCRIPTIVE EXPERIMENTAL Survey Pre-Experimental True Experimental Quasi-Experimental/ CAUSAL- COMPARATIVE Correlational One Shot Case Study Time Series Experiment Pretest-Posttest Control group Observational One-Group Pretest-Posttest Design Solomon Four Group Design Counter Balanced Design Comparative Static Group Comparison Developmental Posttest-Only Control Group Design NonEquivalent Control Group Design Evaluation Factorial Design Ex-Post Facto Trend Analysis Types of Quantitative Research Designs  Descriptive Research Design ►Used as a pre-cursor to other research designs ►Can yield a large amount of data that lead to important recommendations and detailed analysis. Descriptive Research Design ►Cannot be used to discover a definitive answer or to disprove a hypothesis ►Heavily dependent on instrumentation for measurement and observation  Correlational Observational Comparative Survey Studies Studies Studies Developmental Evaluation Trend Studies Studies Analysis TYPES OF DESCRIPTIVE RESEARCH DESIGN ►Attempts to establish norms or standards based on a wide class of survey data. ►Local control and control of extraneous variables are dependent on the limitations of your specific questions Survey Research ►Estimates the extent to which different variables are related to one another in the population of interest Correlational Studies ►Requires comparing the occurrence or nature of two or more variables on samples or subjects within the same period/time Comparative Studies ►Less reliable in terms of actual hypothesis testing ►Conclusions drawn from descriptive designs are at best tentative General Considerations in Descriptive Research  Causal-Comparative ► Used to measure what impact a specific change will have on existing norms and assumptions ► Tries to establish a causal link between variables and eliminates other possibilities. ► Also called QUASI-EXPERIMENTAL Causal-Comparative ►Without random assignment of groups ►The possibility always exists that, by sheer coincidence, two unrelated events appear to be related. ►Causality can only be inferred, never proven.  Control Intact Groups Experimental/ Treatment ►Used when the independent variable (cause) cannot be manipulated because the dependent (effect) has already occurred Ex-Post Facto ► To determine the influence of a variable introduced only after a series of initial observations and only where one group is available O1 O2 O3 X O4 O5 O6 O1,O2,O3 - Pre-observations for Experimental Group O4,O5,O6 - Post-observations for Experimental Group X = Treatment/Intervention Time Series Experiment ► Also called rotation design ► Involves exchange of two or more treatments taken by subjects during the experiment GROUP Day 1 Day 2 Day 3 A X1O2 X3O2 X2O2 B X2O2 X1O2 X3O2 C X3O2 X2O2 X1O2 Counter Balanced Design ► Subjects are not randomly assigned to the different group O1 X O2 O3 O4 O1 = pretest for experimental group O2 = posttest for experimental group O3 = pretest for control group O4 = posttest for control group X = Treatment Nonequivalent Control Group Design EXPERIMENTAL RESEARCH DESIGN Approach provides the highest level of evidence for single studies Allows the researcher to control the situation Random assignment of individuals/units  EXPERIMENTAL RESEARCH DESIGN The design is artificial, and results may not generalize well to the real world Can be costly if special equipment or facilities are needed. Some research problems cannot be studied using an experiment because of ethical or technical reasons Groups with Randomization Control (Negative, Positive) Experimental/Treatment  TYPE TRUE QUASI-EXPERIMENTS EXPERIMENTS LOCATION Laboratory Common Unusual Field Less Common Common TRUE EXPERIMENTAL VS. QUASI-EXPERIMENTAL PRE-EXPERIMENTAL STUDIES Conducted before Usually avoided due the actual lack of experimentation to randomization see if it influences a small group ► The least reliable of all experimental approaches. G: X_ _ _ _ O ONE SHOT CASE STUDY ► An approach that provides a measure of change but can provide no conclusive results ► Involves one group which is pretested, exposed to a treatment, and then post-tested. G: O1 _ _ _ _ _ X_ _ _ _ _O2 ONE GROUP PRETEST- POSTTEST DESIGN ► Involves atleast two groups, one group is exposed to the treatment and both groups posttested. G1 is the experimental group while G2 is the control group. G1: X_ _ _ _ _ O2 G2: _ _ _ _ _ _ O2 STATIC GROUP COMPARISON TRUE EXPERIMENTAL STUDIES Require a control group and random assignment of subjects to groups The dependent variable is the effect of the manipulation of the independent variable ► Involves at least two group ► One group is exposed to a treatment. Both groups are posttested POSTTEST-ONLY CONTROL GROUP DESIGN ► This involves at least two groups, which are both pretested ► One group is not exposed to a treatment (control group) which the other group is exposed to a treatment (experimental group). Both are posttested. PRETEST-POSTTEST WITH CONTROL GROUP ► With the greatest validity ► It requires four groups. Groups 1 and 2 are pretested. Groups 3 and 4 are not pretested. Groups 1 and 3 are exposed to treatment (experimental), while groups 2 and 4 are not exposed to treatment (control groups). All four groups are posttested. SOLOMON FOUR GROUP DESIGN ► Thisdesign allows the researcher to investigate the interactions of more than one independent variables and their effects on the dependent variable. TYPES OF COFFEE DRINKERS R1 R2 R3 TREATMENT Coffee only Coffee with sugar Coffee with sugar and cream X1 X2 FACTORIAL DESIGN  ► Random assignment of individuals in groups ► Random assignment of treatments to groups Completely Randomized Design Completely Randomized Design Mouse A Mouse I Mouse B Mouse D Mouse E Mouse C Mouse H Mouse F Mouse G Completely Randomized Design Treatment B Control Treatment A Mouse A Mouse I Mouse B Mouse D Mouse E Mouse C Mouse H Mouse F Mouse G  ► Usually for agricultural experiments ► Similar experimental units are grouped into blocks Randomized Complete Block Design Randomized Complete Block Design Treatment A Treatment B Treatment C 24 pots with 3 plants Treatment D each Treatment E Control (F)  Randomized Complete Block Design Blocks Treatments A B C D E CTRL 1 Pot 1 Pot 1 Pot 1 Pot 1 Pot 1 Pot 1 2 Pot 2 Pot 2 Pot 2 Pot 2 Pot 2 Pot 2 3 Pot 3 Pot 3 Pot 3 Pot 3 Pot 3 Pot 3 4 Pot 4 Pot 4 Pot 4 Pot 4 Pot 4 Pot 4 Randomized Complete Block Design Pot 1A Pot 1B Pot 1C Pot 1D Pot 1E Pot 1F Pot 2A Pot 2B Pot 2C Pot 2D Pot 2E Pot 2F Pot 3A Pot 3B Pot 3C Pot 3D Pot 3E Pot 3F Pot 4A Pot 4B Pot 4C Pot 4D Pot 4E Pot 4F Randomized Complete Block Design Pot 1A Pot 2B Pot 1C Pot 3D Pot 2E Pot 4F Pot 2A Pot 3B Pot 4C Pot 2D Pot 4E Pot 1F Pot 3A Pot 4B Pot 2C Pot 4D Pot 1E Pot 3F Pot 4A Pot 1B Pot 3C Pot 1D Pot 3E Pot 2F  ► Briefly state the overall design of the research (qualitative or quantitative). Then, state the specific design being observed in the study. ► Provide 1-2 sentences conceptual definition with in-text citation. ► Lastly, explain how the selected research design will be utilized in the context of your research. How to write Research Design in the Methodology (Chapter 3)?  Identify the Research Design Scenario 2: In this study the researchers measured the initial weight of the pears and subjected the pears to initial test of colors to gather pretest data. After the storage period of 26 days, the researchers then determined the final weight of the pears and subjected them to final test of colors to obtain posttest data. The data from the experimental data was compared to the controlled set up as well. PRETEST-POSTTEST WITH CONTROL GROUP  Identify the Research Design Scenario 3: In this field there is a slope that causes a fertility gradient. The field is more fertile at the bottom than at the top. Different colors represent different treatments; each horizontal row represents a block. There are 4 blocks [I-IV] and 4 treatments [A (red); B (blue); C (yellow); D (green)]. Randomized Complete Block Design SAMPLING DESIGN POPULATION SAMPLE Terminologies Population Target Population Sample Subject  Sampling ✔ Process of choosing a representative portion of a population Advantages Can be thoroughly Fewer errors Less cost studied Less time Less effort Probability Sample Done according to some chance mechanism Have equal chance of being selected Non-Probability Sample Selected by judgment Do not have an equal chance of being selected Probability Sampling Simple Stratified Systematic Random Random sampling Sampling Sampling Cluster Multistage Sampling Sampling Simple Random Sampling ►Each unit is given a chance of being included in the sample Lottery Table of Random #s Systematic Physical mixing Tables, calculators, Employs a computers systematic list Alphabetical or Chronological Simple Random Sampling Systematic Sampling ❑ Uses a formula to decide the elements to be included in the: n=N/k ❑ Where n=nth element in the population; N is the population and k is the sample size. Stratified Random Sampling Divides the population into groups Selecting a simple random sample for each of the groups (strata) into which the population has been subdivided. POPULATION  GROUP 1 GROUP 2 GROUP 3 GROUP 4  SAMPLE GROUP SAMPLE 2 GROUP SAMPLE 3 GROUP SAMPLE 4  Random 25-34 years old sample Example: Marathon 35 – 44 years Random Runners old sample 45++ years Random olds Sample  STRATA Example: 25-34 years Random old sample Marathon 35 – 44 years Random Runners old sample 45++ years Random olds Sample Cluster Sampling ❑ Divides the population in clusters (randomly selected) then obtaining simple random samples in each cluster UN TROIS DEUX QUATRE  ONE Random Sample TWO Random Sample Population THREE Random sample FOUR Random sample Example:  Multistage Sampling ❑ Two or more rounds of different random sampling methods Stratified Cluster Convenience Purposive Accidental Quota Snowball NON-PROBABILITY SAMPLING  Based on choosing individuals as samples according to the purpose of the researcher. Purposive Sampling Things are chosen as samples because of the good evidence that they are representatives of the population.  Alternative Feeds for Ducks Antimicrobial Activity of Mango Peel Example Social Development of Refugee Children Evaluating the Grade 10 Biology Curriculum Convenience Sampling ❑ Also called incidental sampling ❑ Applied to samples which are taken because they are the most available ❑ Simply takes the nearest objects as the sample till the sample reaches the desired amount/size. Example ► Initial Reactions on the movie “Black Widow” Quota Sampling Done to insure the Deliberately sets a inclusion of a proportion within particular segment the sample of a population Example ►Attitudes of Members of Different Religions towards Divorce Bill  No Includes system anyone Accidental Sampling Snowball Sampling ❑ Uses network ❑ One respondent will refer another respondents Keep in mind… For natural sciences research, sampling procedures are taken in a different context. The sampling techniques described here are typically used in social science research, where human subjects are studied, and participants are chosen at random. Thus, population and sampling is not reflected on your Methodology (Chapter 3).