Presentation on theme: "Designs that allow testing of hypotheses. Describe pre-experimental, experimental and quasi-experimental research designs. Explain the types of conclusions."— Presentation transcript:
Describe pre-experimental, experimental and quasi-experimental research designs. Explain the types of conclusions you can draw from data obtained in at least three different types of research designs described. Read a research article and describe the research design Describe the research design for your research project.
Qualitative – construct meaning Descriptive – quantitative descriptive stats ◦ identifying characteristics of an observed phenomenon or exploring possible associations among two or ore phenomena. Examines a situation as it is Correlations Surveys, questionnaires, structured interviews, checklists NO causality
Variables ◦ Independent ◦ Dependent Experimental Designs ◦ Experimental ◦ Quasi-experimental ◦ Post Facto Designs
Does not show cause & effect Independent “variable” does not vary Experimental/control groups not equivalent or randomly selected Usefulness: Forming tentative hypotheses more controlled studies
One-shot experimental case study One Group Pre/Post test Static Group Comparison Time GroupTime Group 1Treatment (Tx)Observation (obs) GroupTime Group 1ObsTxObs GroupTime Group 1TxObs Group 2 ---Obs
Random assignment of subjects causality Pre/Post test Control Group Solomon Four-Group Design - ◦ Does the pre-test impact the post test? GroupTime Randomly Assigned 1ObsTxObs 2 GroupTime Randomly Assigned 1ObsTxObs 2 ---Obs 3---TxObs 4--- Obs
Posttest Only Control Group Within Subjects Design – i.e. one exam (Obs), ◦ Tx1 – some content taught with case studies, some content taught by pure lecture GroupTime Randomly Assigned 1TxObs 2---Obs GroupTime 1Tx 1 Obs Tx 2 Obs
Randomized subjects impossible or not practical Don’t control for all confounding variables - need to account for them in discussion Nonrandomized Control Group Pre/post See others on table GroupTime 1ObsTxObs 2 ---Obs
After the fact Results do not allow causality conclusions i.e. study of students who have failed a course once and are retaking it. GroupTime Prior Event(s) 1Exposure (Exp)Obs 2---Obs
A researcher studies the effects of two different kinds of note-taking training (one of which is placebo) on the kinds of notes that college students take. Her sample consists of students enrolled in two sections of an undergraduate course in educational psychology; with the flip of a coin, she randomly determines which section will be the treatment group and which will be the control group. She analyzes the content of students’ class notes both before and after the training, making the prediction that the two groups’ notes will be similar before the training but qualitatively different after the training. Quasi-experimental: Nonrandomized control group pre/post test (Design 8)
Even with control groups, not all research designs allow you to draw conclusions about causality. Pick a research design that allows you to make the types of conclusions about the data that you are interested in making.