Essays.club - Get Free Essays and Term Papers
Search

Psychology 240 - Comp Lab 2: Descriptives, Z-Scores, Power

Autor:   •  November 27, 2017  •  992 Words (4 Pages)  •  607 Views

Page 1 of 4

...

Run Descriptive Statistics on Z-scores

- Get the means and ranges for the two Z-score variables. What are the means? On average, how many standard deviations away from the mean are the number of sick days? What about for SLEs? Are you surprised by these values? (2 points)

The means are so close to zero that it can’t determine the mean. The minimum standard deviation for SLEs is 1.39 and the max is 2.38. The minimum standard deviation for sick days is 1.40 and the maximum 2.49. The average standard deviation is 1. The z scores are the number of standard deviations away from the mean which explains why the mean here would be 0 therefore I was not surprised by the values.

- What is the range of z-scores for sick days? What about for SLEs? (1 point)

The range of z-scores for sick days is 3.88 and for SLEs 3.77

Use GPOWER to calculate power

Post-hoc power analysis: In this data sample, we had 24 participants and want to estimate our power to detect a medium effect (f2 = .15) with the 2 predictors.

- Open GPOWER, click on “Tests” → “Correlation and Regression” → “Linear Multiple Regression: Fixed Model, R² Deviation from Zero.”

- Change type of power analysis to “Post Hoc:…” (middle of screen).

- Input the relevant parameters, including effect size (f2 = .15), alpha level (α = .05), sample size (24), and number of predictors (2).

- Post-hoc power analyses determine what your power will be given certain parameters like sample size.

What was our power for these analyses? (i.e. what was our chances of finding a significant effect of this size, if it existed; 1 point)

Our power is .33 or 33%.

If you were actually running this experiment, what would you do to improve these chances? (1 point)

I would have had more participants.

Plot power as a function of sample size

Click “X-Y plot for a range of values” (towards bottom right) and then click “draw plot” (towards bottom right of new window).

- About how much would your power increase if we had 50 participants? (1 point)

It would be at .66 or 66% so it would have increased by .33 or 33%.

- About how many participants would you need to have power = .80? (1 point)

Approximately 69.

...

Download:   txt (6.1 Kb)   pdf (73.9 Kb)   docx (570.1 Kb)  
Continue for 3 more pages »
Only available on Essays.club