7126/6667 Survey Research & Design in Psychology Semester 1, 2011, University of Canberra, ACT, Australia James T. Neill Home page: http://ucspace.canberra.edu.au/display/7126 Lecture page: http://ucspace.canberra.edu.au/display/7126/Lecture+-+Review Notes page: http://en.wikiversity.org/wiki/Survey_research_methods_and_design_in_psychology Image name: Recycle symbol Taiwan.svg Image source: http://commons.wikimedia.org/wiki/File:Recycle_symbol_Taiwan.svg Image author:Bryan Derksen http://commons.wikimedia.org/wiki/User:Bryan_Derksen License : Public domain Description: Reviews this semester-long (150-hour), third year undergraduate psychology research unit which focused on survey research methods and survey design. This lecture emphasises the second-half of the unit's content on MLR, ANOVA, significance testing, power, and effect size, as well as providing advice about the lab report and final exam assessment items.
Aims & outcomes <ul><li>Knowledge & skills for conducting ethical , well-designed , survey-based research in psychology. </li></ul>How confident are you that could conduct a good quality survey-based research study? For 4 th year Honours? In the work-place?
Aims & outcomes <ul><li>Theory & practice of survey-based research, incl.: </li></ul><ul><ul><li>Research questions / hypotheses
Visualisation of data <ul><li>Aids interpretation of descriptives and tests of differences or relationships.
Univariate : </li><ul><li>histogram, bar graph, error-bar graph </li></ul><li>Bivariate : </li><ul><li>scatterplot, clustered bar graph </li></ul><li>Multivariate : </li><ul><li>Venn diagrams, multiple line graph, 3-d scatterplot </li></ul></ul>
Software for data visualisation (graphing) <ul><li>Statistical packages </li></ul><ul><ul><li>e.g., SPSS Graphs or via Analyses </li></ul></ul><ul><li>Spreadsheet packages </li></ul><ul><ul><li>e.g., MS Excel </li></ul></ul><ul><li>Word-processors </li></ul><ul><ul><li>e.g., MS Word – Insert – Object – Micrograph Graph Chart </li></ul></ul>
Multiple linear regression <ul><li>L inear regression formula </li></ul>Y hat = ax + b Y = ax + b + e <ul><li>Proportion of variance in a DV explained by one or more IVs </li><ul><li>R , R 2 , Adjusted R 2
<ul><li>Overall hypothesis : </li><ul><li>(Null) That the IVs do not explain variance in the DV (i.e., that R is 0) </li></ul><li>+ one hypothesis per predictor : </li><ul><li>(Null) That each IV is not a significant predictor of variance in the DV (i.e., that t for each predictor is less than the critical value) </li></ul></ul>Multiple linear regression
<ul><li>Consider/interpret : </li><ul><li>Direction of each predictor
Significance testing Significance testing has dominated psychology, but is problematic, mainly because: <ul><ul><li>Results are dichotomous (sig. or not), which doesn't help us to understand the size of effect.
Sig. test results are influenced by power esp. if particularly high or low. </li></ul></ul>
Power & effect sizes Power and effect sizes have been neglected. Therefore: <ul><ul><li>Calculate the power of studies (prospectively &/or retrospectively)
Report ESs & CIs to complement inferential statistics </li></ul></ul><ul><ul><ul><li>r, r 2 , R , R 2
Publication bias & academic integrity Publication bias (low power; favouritism of sig. findings; funnel plots) Academic integrity - “Integrity is doing the right thing, especially when no one is watching”.
Lab report - Discussion <ul><li>Provide insight about the results
Draw conclusions about the RQs & hypotheses in light of the results.
Discuss key strengths & limitations of the study. ( Balanced criticism )
Draw out implications and recommendations </li></ul>
Lab report - Discussion <ul><li>Offer specific, practical recommendations e.g., </li></ul><ul><ul><li>Theory : What are the implications for the theory/rationale upon which the study was based?
Methods : How could the research design (e.g., instrumentation) be improved?
Practice : Implications for students and universities e.g., for improving satisfaction? </li></ul></ul>
Lab report – Appendices <ul><li>Optional : Include appendices where relevant and referred to in the body text. Appendices may not be consulted by a reader, so if its important make sure key content is covered in the text .
Use for content which would break the flow , but which is relevant to understanding the study e.g., the EFA correlation matrix. </li></ul>
Lab report – Appendices <ul><li>APA style not necessary.
Use headings (e.g., Appendix A, B, C etc.) and possibly titles e.g., </li></ul>Appendix A: Bivariate correlations amongst the university student satisfaction items