Part I Design Issues:
Explain both the similarity and difference between the two key terms in each option. In your answer of about 30-50 words, explain:
- How the two terms are related, and
- How they are different.
| Terms | Your Answer |
| Internal and External Validity | a b |
| Between Subjects and Within Subjects Design | a b |
| Type I (alpha) and Type 2(beta) error | a b |
| A t test and Oneway Analysis of Variance | a b |
| Spearman and Pearson Coefficients | a b |
| Bonferroni and LSD Post Hoc Tests | a b |
| Experimental and Non-experimental Designs | a b |
| Positively and Negatively Skewed Distributions | a b |
| Normal and Bimodal Distributions | a b |
| Random Sample and Stratified (or proportional) Sample | a b |
Part II Understanding Research:
I will describe a research design. Your job is to answer the following 7 questions about it. Each question is worth 4 points, for a total of 28 points. Answers will usually be at most a sentence long: sometimes only a word or two is needed.
Researcher VG wanted to find a way to improve student study habits for the Introduction to Psychology course. VG studied the literature and decided to use four different study methods that were described in it.
Method A: unstructured study
Method B: structured study with outline
Method C: structured study with outline and discussion groups
Method D: structured study with outline, discussion groups, and practice tests.
Students studied in the specified way all term, and then took the same final exam at the end.
| Question | Your Answer |
| Approximately how many participants (in total) should VG use? Be realistic (VG does not have years to complete the research); also be statistically wise. | |
| If we wanted to describe VG’s study in terms of a Control Group, which Method (A, B, C, or D) would be the control group? Why? | |
| Why might VG want to use SPSS to make sure that the distribution of scores in the final exam was normal? Which SPSS procedure could VG use to do this? | |
| Which inferential procedure/s should VG use to compare the group averages? | |
| What would the null hypothesis of the statistical test in (4) be? The alternative hypothesis? | |
| VG had a favorite method – Method D. VG believed D would be better than all the other methods. How could this expectation be tested statistically? | |
| How might VG use the technique of random assignment to groups in this research? |
Part III Reading SPSS output,
This SPSS output is the result that VG got from data analysis for the research in the previous question. Answer the eleven questions below, 4 points each. Use 30 words or less for each answer. Sometimes only one or two words are enough.
| Descriptives | |||||
| finalexamscore | |||||
| N | Mean | Std. Deviation | Std. Error | ||
| A | 8 | 66.7500 | 13.41375 | 4.74248 | |
| B | 9 | 76.6667 | 4.30116 | 1.43372 | |
| C | 7 | 79.8571 | 5.39841 | 2.04041 | |
| D | 8 | 88.2500 | 5.14782 | 1.82003 | |
| Total | 32 | 77.7813 | 10.86496 | 1.92067 |
| Tests of Homogeneity of Variances | |||||
| Levene Statistic | df1 | df2 | Sig. | ||
| finalexamscore | Based on Mean | 2.832 | 3 | 28 | .056 |
| Based on Median | 2.650 | 3 | 28 | .068 | |
| Based on Median and with adjusted df | 2.650 | 3 | 14.239 | .089 | |
| Based on trimmed mean | 2.744 | 3 | 28 | .062 |
| ANOVA | |||||
| finalexamscore | |||||
| Sum of Squares | df | Mean Square | F | Sig. | |
| Between Groups | 1891.612 | 3 | 630.537 | 9.987 | .000 |
| Within Groups | 1767.857 | 28 | 63.138 | ||
| Total | 3659.469 | 31 |
| Multiple Comparisons | ||||||
| Dependent Variable: finalexamscore | ||||||
| LSD | ||||||
| (I) method | (J) method | Mean Difference (I-J) | Std. Error | Sig. | 95% Confidence Interval | |
| Lower Bound | Upper Bound | |||||
| A | B | -9.91667* | 3.86103 | .016 | -17.8256 | -2.0077 |
| C | -13.10714* | 4.11241 | .004 | -21.5310 | -4.6833 | |
| D | -21.50000* | 3.97296 | .000 | -29.6382 | -13.3618 | |
| B | A | 9.91667* | 3.86103 | .016 | 2.0077 | 17.8256 |
| C | -3.19048 | 4.00437 | .432 | -11.3931 | 5.0121 | |
| D | -11.58333* | 3.86103 | .006 | -19.4923 | -3.6744 | |
| C | A | 13.10714* | 4.11241 | .004 | 4.6833 | 21.5310 |
| B | 3.19048 | 4.00437 | .432 | -5.0121 | 11.3931 | |
| D | -8.39286 | 4.11241 | .051 | -16.8167 | .0310 | |
| D | A | 21.50000* | 3.97296 | .000 | 13.3618 | 29.6382 |
| B | 11.58333* | 3.86103 | .006 | 3.6744 | 19.4923 | |
| C | 8.39286 | 4.11241 | .051 | -.0310 | 16.8167 | |
| *. The mean difference is significant at the 0.05 level. |
| Question | Your Answer |
| How many participants used study Method C in this research? | |
| Which of the four groups (A, B, C, or D) had the widest spread of scores (greatest variability)? How do you know? | |
| Was the assumption of homogeneity of variance satisfied? How do you know? | |
| Was the main null hypothesis of the research rejected? How do you know? | |
| The F value of 9.987 comes from dividing one number in the ANOVA table by another (it is a ratio). What are the two numbers? | _____________ is divided by _______________ |
| In comparison to a Bonferroni, was the post hoc test employed likely to find and report more or fewer significant differences? How do you know? | |
| Which two of the study Methods were most similar to each other in terms of final exam mean? | |
| According to the post-hoc test, were the two means in your answer to the previous question (7) significantly different? | |
| Was VG’s expectation (that Method D would provide the best results) supported by the data and analysis? How do you know? | |
| How many degrees of freedom does the F test have? | |
| What was the average score on the final exam for Group A? |