Paired samples t test interpretation help (included image of SPSS output)

#1
Hi,

My undergraduate dissertation is on change in cyberchondria between 2019 and 2021, in relation to COVID-19. I computed the variables CSS_THEN, CSS_NOW then calculated CSS_CHANGE ** (CSS = Cyberchondria severity scale)*

This is my first draft of writing it up:

A paired-samples t-test was conducted to compare the change of cyberchondria severity between 2019 (then) and 2021 (now). There was a significant difference in the scores for cyberchondria severity in 2019 (M= 2.30 , SD= 0.67) and cyberchondria severity in 2021 (M= 2.52, SD= 0.69); t(49) = -3.39, p = .001. These results suggest that cyberchondria severity has increased since 2019.

Could anyone help me expand on this or improve it? I'm a beginner to stats.

Thankyou
 

Attachments

#2
Hi,

My undergraduate dissertation is on change in cyberchondria between 2019 and 2021, in relation to COVID-19. I computed the variables CSS_THEN, CSS_NOW then calculated CSS_CHANGE ** (CSS = Cyberchondria severity scale)*

This is my first draft of writing it up:

A paired-samples t-test was conducted to compare the change of cyberchondria severity between 2019 (then) and 2021 (now). There was a significant difference in the scores for cyberchondria severity in 2019 (M= 2.30 , SD= 0.67) and cyberchondria severity in 2021 (M= 2.52, SD= 0.69); t(49) = -3.39, p = .001. These results suggest that cyberchondria severity has increased since 2019.

Could anyone help me expand on this or improve it? I'm a beginner to stats.

Thankyou

Difference is significant (p = 0.001) and there is increase in mean_css (0.2167).

Did your sample change from 2019 to 2021 or are they the same? If they are the same sample getting the same scale than you should the paired-samples-test. If not independent samples t-test is more appropriate.
 
#3
Thanks! Yeah same sample. We just gave the questionnaire twice, once for how they felt in 2019, and once for how they feel now (definitely a limitation for my discussion haha)
 
#4
Okay then paired-sample t-test is fine. Why do you think this is a limitation. Actually measuring the result from an unchanged sample can be a strength of your study. If the samples were different your would be less reliable at showing the change in mean_css. Different sample would introduce new individuals with different qualities which could introduce confounding factors.