1 2 Research Methodology Week 4 Characteristics of Observations 1. Most important know what is being observed. 2. Assign behaviors to categories. 3. Know how to Measure. 4. Degree of Observer inference. 3 Research involves trying to determine the relationship between two or more variables. The researcher asks: What is the effect of X on Y? How is X related to Y? 4 5 6 Some examples might be: IQ Anxiety Academic success Gender Income Recidivism rates In order to study variables, we must define them clearly, and have some way to measure them. In research we call this the process of creating operational definitions. For example: Treatment outcomes may be measured by changes in behavior (going to the store alone) or changes on a scale (such as an anxiety scale). Independent A variable that is manipulated to examine its impact on a dependent variable. Manipulated by the researcher. Independent of the outcome. 1
Presumed to cause, effect or influence the outcome. 7 Dependent A variable that indicates whether the treatment or manipulation of the independent variable had an effect. Dependent on the independent variable(s). Depends on how the independent variable(s) are managed or manipulated. 8 9 10 Extraneous A variable that is related to the dependent or independent variable that is not part of the experiment. Variance Differences among phenomena - Without differences and variation there is no way to determine relations among variables. In research, means of different experimental groups are compared to study relations. Error Variance Error Variance Is the fluctuation or varying of measures due to chance (random variance). Variation in measures due to the usually small and self-compensating fluctuations of measures. 11 12 13 Sampling taking a portion of the population and considering it to be representative. Within the Individual General Level of ability Test taking skills Ability to understand instructions Specific Ability related to the trait being assessed. Test-taking skills specific to the test. Temporary Health, fatigue, emotional strain, motivation. 2
Outside the Individual Test Administration Factors Conditions of testing Interaction between examiner and test taker. Bias in grading. Rater Bias. 14 Interpretation of data and reports Research report (different formats) 1. Intro or Rational 2. Review of Literature 3. Method (research design) 4. Results and discussion 5. Summary 15 16 17 18 Research Biases -Placebo s -Researcher Bias Deliberate or inadvertent bias in which data are misanalyzed or participants are treated differently over and above any planned differences in treatment. -Single / Double Blind Studies -Hawthorne Effect Refers to conditions under which performance in an experiment is affected by the knowledge of participants that they are in an experiment. Rater Bias Ethical Considerations Protection from harm Minimize the risk to participants in the study, including, if needed, follow-up contact to resolve potential negative consequences associated with their participation. Debriefing Is an ethically required process where the researcher explains the nature and purpose of his/her research to the participants. Research Design is a plan, structure, strategy of investigation conceived to obtain answers to research questions and to control variance. Good research does not give us answers 3
only new questions! 19 20 21 22 23 24 Research Designs True Experimental To investigate possible cause-and-effect relationships by exploring one or more experimental groups to one or more treatment conditions and comparing the results to one or more control groups not receiving the treatment (random assignment being essential). Treatment / Control Group Treatment Group - Is the set of participants in an experiment that receives the treatment/independent variable. Control Group - Is the set of participants in an experiment that does not receive the treatment/independent variable. Data from the control group are compared against the experimental group to evaluate the potential effect of an independent variable. Basic Research Designs Historical To reconstruct the past objectively and accurately. Descriptive To describe systematically a situation or area of interest factually and accurately. Developmental To investigate patterns and sequences of growth and/or change as a function of time. Case and Field To study intensely the background, current status, and environmental interactions of a given social unit: an individual, group, institution, or community. Basic Research Designs Correlational To compare variables to determine a relationship of + /-. Ex-Post Facto To investigate possible cause-and-effect relationships of past events. Quasi Experimental Is a type of experimental research design that contains a variable that cannot be directly manipulated or controlled be the researcher. Survey Uses questionnaires to collect information from a large number of individuals. Pilot Study Refers to preliminary research that is conducted on a small scale (i.e., with a small number of participants) in order to evaluate the appropriateness of methods (e.g., choice of variables and their selected levels). Research Hypothesis A hypothesis converts the question posed by the research problem into a declarative statement that predicts an expected outcome. 4
A hypothesis is an assumptive statement about the relationships between two or more variables that suggests an answer to a research question. 25 26 27 Research Hypothesis A hypothesis should be expressed in clear, simple, and concise terms. The person reading a hypothesis should be able to determine: The variables of the hypothesis. The population being studied. The predicted outcome of the hypothesis. Research and Null Hypotheses Research What you are testing (you are saying there is a difference). Null Population (says there is no difference, if so it is due to chance or error. Null Results occurs when a dependent measure is not affected by manipulations of the levels of one or more independent variables. -You only test the null hypothesis to: -Reject or -Fail to reject 28 29 30 31 Type I Error Rejecting the null hypothesis when it is true. (deciding an ineffective treatment is effective). Type II Error Accepting the null hypothesis when it is false. (failing to identify that an effective treatment is actually effective). As the probability of making a Type I error decreases, the probability of making a Type II error increases. That is as you try to make sure that you won t make false claims about a drug by making α (sigma) small, you increase the probability of overlooking a treatment that actually works. Scientific Proof Science does not Prove anything It only gives evidence to except or not except a given hypothesis Research Overview Topic Update 5