Section A: Factorial ANOVA and Interactions

Section B: Hierarchical Linear Regression

Section C: Logistic Regression

The assignment (portfolio) is divided into different sections and assesses your understanding of quantitative analyses (e.g. ANOVA & regression techniques). You should try to answer all questions – some of which can be answered based on lecture material and others require you to engage in additional reading. Please support your answers with references wherever possible. When presenting your portfolio please label the three sections as indicated above and do not exceed any word counts that are provided. All answers must be presented in a narrative – do not use bullet points otherwise marks may not be awarded for that question/section. You will need to organise your time effectively in completing this work to allow you to meet the deadline – therefore please start work on the portfolio as soon as possible. The questions will be easier to address once you have engaged with the relevant lectures and workshops and additional reading, and you can use the workshops to raise questions that you may have about aspects of the portfolio.

Section A : Factorial ANOVA and Interactions

1. Inspect Table 1. The data are taken from an experiment examining visual search performance in which participants searched for a target among distractors. The within-participants factors were trial type (the target was either present or absent from the display) and target type (digits or letters) and the between-participants factor was working memory group. Are the patterns of means suggestive of any main effects? Explain your answer. [60 words]
   Table 1
   Means (ms) for target present and absent trials for letters and digits in low and high working memory groups

   Trial Types	Letters	Digits
   	Low Working Memory
   Target Present Trials	1528	1664
   Target Absent Trials	2044	2480
   	High Working Memory
   Target Present Trials	1222	1201
   Target Absent Trials	1523	1502

2. The data on Blackboard are taken from an experiment that examined visual search performance in men and women. Participants searched for a target item among distractors in displays with a small set size (12 items in the display) and a large set size (24 items in the display). The target was present in the display on 50% of trials but absent on the remaining trials. The dependent variable was search times recorded in milliseconds (ms). Using the descriptive data available, produce an APA table of means and standard deviations. Remember to be consistent with decimal places and to include an appropriate table number and title.
3. Using the SPSS data output on Blackboard, report the results of the study (all main effects and interactions) in a narrative. You should include F ratios, MSEs, p values and effect sizes. State the exact p value unless this is < .001. You should refer to all effects by variable name (i.e. the main effect of set size; the interaction between set size and trial type) and ensure that you make it clear whether each effect is significant. [160 words]
4. Using the SPSS data file on Blackboard run the appropriate post hoc tests to examine the trial type x set size interaction and report the results of the post hocs in an appropriate narrative. [80 words]
5. Summarise the trial type x set size interaction using the descriptive information where necessary to support your explanation. [120 words]
6. The post hoc tests for the trial type x gender interaction have been conducted for you (see SPSS output on Blackboard). With reference to this information and the bonferroni correction, explain what the interaction shows. Use descriptive information where necessary to support your explanation. [120 words]
7. Explain the difference between cyclic and digraph counterbalancing and comment briefly on why they should be considered when designing experiments. [50 words]
8. Summarise the data shown in Figures 1a and 1b (Appendix A) and comment on whether this suggests evidence of an interaction and provide evidence to support your comments. [120 words]

Section B : Hierarchical Linear Regression

1. Comment briefly on the assumptions for hierarchical linear regression [20 words].
2. The SPSS data output on Blackboard shows a hierarchical regression analysis in which the criterion was spelling performance in children. Non-verbal reasoning skills (Raven’s) were entered first, followed by two language measures (TROG & expressive vocabulary) and phoneme awareness (spoonerisms and alliteration) was entered in the final model (see Appendix B for a summary of the tests). Present an APA formatted table to show the B, SE of B and Beta values for the data shown. Remember to number the table and to provide an appropriate table title.
3. Using the SPSS output on Blackboard, report all the results of the multiple regression analysis (see lecture and guidance notes). Provide an interpretation for these results including a statement of what beta values represent and therefore what they show for this particular study. [300 – 350 words]
4. Supporting your answer with references, provide a brief discussion of the influence that residuals play in multiple regression analysis and the extent to which they may compromise the integrity of the results. [150 words].

Section C : Logistic Regression

1. Following the format of the table used in the lecture, present an APA table showing the relevant statistics for the logistic regression analysis (see blackboard for SPSS output). In this analysis, the criterion was spelling group with phoneme awareness (spoonerisms) entered first and then TROG entered next (i.e. hierarchical). The table should be numbered and have an appropriate title.
2. The results from the logistic regression analysis on blackboard are reported for you (see Appendix C). Drawing on these results answer the following questions:

1. Compare the constant model with the models from block 1 and block 2, and comment on the classification rates for all models. What do the classification and hence misclassification rates tell us about the importance of the predictors? [50 words].
2. What conclusion can you draw from the -2LL values when there were no predictors in the model relative to when spoonerisms and then TROG were entered into the model? [50 words].
3. Explain what it means that the block statistics revealed a non-significant effect even though the model was significant overall. What does this allow us to conclude about the predictors? [50 words].
4. Provide a definition of the Exp B value and explain what these values mean for this study. [50 words].
5. How do the assumptions for logistic regression differ to those for hierarchical linear regression? [50 words].
6. When might you use multinomial logistic regression? [30 words].
7. What other values from logistic regression might be reported? Explain your answer.

Price: $39.93

Solution: The downloadable solution consists of 15 pages, 2493 words and 3 charts.
Deliverable: Word Document