Week 7 Research and Program Evaluation.pdf

Week 7: Research and Program Evaluation • 3 common types of variables• Independent : manipulated to create a desired outcome • Dependent: affected by the independent variable • Control: constant throughout the study • 4 Types of Sampling Methods • Simple random sample: each sample has an equal chan...

Week 7: Research and Program Evaluation • 3 common types of variables• Independent : manipulated to create a desired outcome • Dependent: affected by the independent variable • Control: constant throughout the study • 4 Types of Sampling Methods • Simple random sample: each sample has an equal chance of selection • Stratified random sample: samples are grouped then randomly selected • Cluster sample: similar sample groups are researched together • Systematic Random Sample: samples selected by predictable intervals • Hawthorne Effect: placebo effect; knowing you are in a study will skew the results • Novelty Effect: good initial results but later taper as the newness wears off • Experimenter Effect: data gets skewed because of researcher behavior • Halo Effect: researcher’s initial thoughts cause skewed reporting • Factor analysis: designed to reduce a lot of variables down to a smaller list of factors • Meta Analysis: blending of similar study results to review the outcomes • Alpha error: original hypothesis (null) said there was NO difference in the variables BUT research showed there ARE differences thus the null is REJECTED • Beta error: the original hypothesis said there was NO difference in the variables AND the research proved it to be correct, so the null is ACCEPTED. • Type I Error (Alpha Probability) – False Positive • Type II Error (Beta Probability) – False Negative • Null Hypothesis (Ho) assumes that there is NO difference between groups/variables/etc. • Alternative Hypothesis (H1) The Research Hypothesis; the effect observed in the data (the sample) reflects a “real” effect. • Percentage Score = Raw Score • Variance = SD*SD • Range = Biggest Number – Smallest Number • Skew = mean-median; the bigger the difference, the greater the distribution skew • If the mean and median are equal, the distribution is normal • If the mean is greater than the median, the distribution is positive • Strong positive correlation coefficients are as close to 1.00 as possible but aren’t expressed in percentages. • Types of Measurement Scales • Nominal (categories) = lowest level of measurement • Ordinal (order) = order of subjects • Interval (numbers) = equal; most educational research are interval • Ratio = highest level of measurement • Qualitative Research: loos to figure out how behaviors occur through observation • Quantitative Research: looks for relationships that can be measured numerically • Evidence-based research: interventions with scientific research that supports client outcomes • Standard Error of Measurement (SEM) – measures the variability of a person taking the same test several times. 68% chance that the test scores will fall within the SEM provided. Low SEM indicates high score accuracy