Screen Cast videos 12.8 and 12.8.2

Preparation

To prepare for this discussion, follow along with our Screen Cast videos 12.8 and 12.8.2 to make sure you can generate all of the required results.

Required SPSS Videos

    1. Stats 3e Screencast 12.8 (shows how to get a One-Way Anova from the “compare means” menu)
    2. Stats 3e Screencast 12.8.2 (shows how to get the same results but from the GLM (general linear model) menu)
      1. This should be further evidence that Anova is a special case of the general linear model.  You will learn more about that later in our class.
      2. Pay special attention to the mention of the “power” result which is available through the GLM command and which our speaker notes is an estimate of what percent of samples would yield a significant result IF there was a true effect in the population.  This is interesting because in the video example, that percent is about 61, which is relatively low power
    3. Use this NICE video (click here) to see how to calculate EFFECT SIZE for the One-Way Anova (Eta squared, partial Eta Squared).  This will also require that you use the steps in video 12.8.2 as the way to generate your One-Way Anova
Optional papers and videos for interpreting SPSS One Way Anova output tables
  1. [Paper] Analysis of variance (ANOVA) comparing means of more than two groups” (click here)
  2. There are any number of youtube videos for interpreting One-Way Anova tables, so if you want to hear more, go to Youtube and take your pick.  Videos use different versions of SPSS, but the main theory for reading the ouptut is the same.
  3. A power-point presentation (click here) on running and interpreting One-Way Anova in SPSS

Mini-Report Instructions

Welcome to our third mini-report.  Now that you have some practice writing these reports, we will add in tthree new features.

  • The concept of operational definitions (a review from our first few weeks of class)
  • The concept of post-hoc testing which is required for any One-Way Anova that has three or more “levels” of the independent variable
  • Note the measures of effect size and statistical power which come with the output tables. Where are these values and what do they mean?

This week we will use the same mini-report format you learned last week to explore the One-Way Anova. The point of a mini-report format is to build your ability to take a research question from start to finish, from creating the question and phrasing it as a testable hypothesis to choosing the right statistical test, generating the results and sharing those results in an APA formatted mini-report.  These mini-reports follow the layout of a typical journal article so you should be seeing that too.

  • Our goals here are the same as last week except that you will need to come up with a scenario in which you COMPARE AMONG 3 OR MORE GROUP MEANS.
  • Here, you will create a mock study and write up the results as a mini-paper.
  • You may use the scenario you thought up in the first discussion of the week, based on Paul Bloom’s TED.com talk, or you can use a different example.
  • We are going to share these mini-reports with each other as a discussion so everybody will be able to see each others’ examples and therefore have a better idea of how to run statistical tests and write up the results.  If you are stumped on how to write your post, look at what others have written but change the example hypothesis to one which interests you.

To complete this assignment, you are asked to generate a scenario that would require the use of a One-Way Anova, write your research question and matching hypothesis, generate fake/mock data, enter that correctly into SPSS, calculate and explain your results.  It would be best to write your replies in MS word, save them, and paste them into this discussion so you don’t lose any work.

The sections of this mini-report are
  1. Title
  2. (2 points) Brief introduction to the scenario (explain what you are comparing, what the variables are and what your null and alternative hypotheses are.
    1. List the independent variable’s name ____ how many levels it has _____ and what those levels are ______.  See this link if you are stumped. https://www.statisticshowto.datasciencecentral.com/probability-and-statistics/hypothesis-testing/anova/
      1. Provide an operational definition for the independent variable (yes, even categorical independent variables need operational definitions).
    2. List the dependent variable’s name and its operational definition_____.  Is this an interval or ratio scale measurement?
  3. (3 pts) Brief explanation of the methods
    1. explain the sample size, sampling method, and measurement methods you might use to have collected the mock data for your study.  Use about 20-30 data points per group.
    2. explain the descriptive statistics and charts which you might use to first analyze the data and check the sample for outliers and normality
    3. list your choice of alpha level and explain that choice.  Will you use the standard p<.05 level or a different criterion and why?
    4. Explain what post-hoc tests you used and why they were used.
  1. (3 pts) Results [Hint, see: this power-point presentation (click here) on running and interpreting One-Way Anova in SPSS]
    1. List the descriptive statistics for your data.  List any charts which accompany them.
    2. Paste in the Anova results table for your data
    3. Report the descriptive statistics (means and standard deviations) and the inferential statistics (One-Way Anova results) in APA format. http://statistics-help-for-students.com/How_do_I_report_a_1_way_between_subjects_ANOVA_in_APA_style.htm
    4. Explain whether or not you rejected the null hypothesis and why

Use this NICE video (click here) to see how to calculate EFFECT SIZE for the One-Way Anova (Eta squared, partial Eta Squared).  This will also require that you use the steps in video 12.8.2 as the way to generate your One-Way Anova