Download presentation

Presentation is loading. Please wait.

Published byRita Kearley Modified over 2 years ago

1
Dependent t-tests

2
Use when: You are examining differences between groups (times) You are comparing two groups (time points) Each participant was tested more than once

3
Time1score – Time2score ___________________________________ Spread t is increased (likely significant) when: ◦ Differences between two time points are large ◦ When spread is very small

4
Requires fewer subjects Ideal for studying changes over time Controls for the impact of individual differences But sensitive to order effects and other changes over time

5
t = ΣD nΣD 2 – (ΣD) 2 n -1 ΣD = sum of differences between scores n = number of pairs of scores

6
Study: ◦ Effects of “study drugs” on performance ◦ Given study drugs before a memory test (Adderall?) (75, 80, 82, 90, 80) ◦ Not given study drugs before a 2 nd memory test (85, 90, 95, 95, 90)

7
1. State hypotheses ◦ Null hypothesis: no difference between scores on drugs versus scores off drugs μ drug = μ nodrug ◦ Research hypothesis: difference between scores on drugs versus scores off drugs X drug ≠ X nodrug

8
t = ΣD nΣD 2 – (ΣD) 2 n -1 ΣD = sum of differences between scores n = number of pairs of scores

9
6. Determine whether the statistic exceeds the critical value ◦ -7.45 > 2.78 (and 4.60) ◦ So it exceeds the critical value 7. If over the critical value, reject the null & conclude that there was a difference in test scores after taking performance drugs versus not taking them

10
In results ◦ Participants scored significantly higher on tests when they did not take performance enhancing drugs (M = 91.00, SD = 4.18) than when they did take the drugs (M = 81.40, SD = 5.46), t(4) = -7.45, p <.05 (can say p <.01).

Similar presentations

© 2017 SlidePlayer.com Inc.

All rights reserved.

Ads by Google