Hello, I am writing my master dissertation at the moment, and your videos are of great help! Thank you very much :) I do have a question though; I am sorry if the question is strange; I'm new on SPSS. I am doing both dependent and independent t-tests for my data, so I've watched your videos on each test. I noticed that you have conducted the normal distribution test differently in each video. Is it enough to conduct normal distribution test just one time, and then conduct paired t-test and independent t-test?
Hi, thanks for your question and glad to hear the videos have been helpful! It wouldn't be possible to check the normality assumption for both tests at once as the independent t-test assumes that the data are normally distributed within each group whereas the paired t-test assumes that the differences between the two time-points are normally distributed.
Hi there! Your videos have been so helpful! I have a quick question I am currently analysing data for SD-WISE 7 item scale - I placed my pre and post-intervention scores into Shapiro-Wilk and It comes as normally disturbed (p = .032). Following this I would use a paired sample t-test? Just so confused haha any help would be appreciated :)
Hi, glad the videos have been helpful! If the p value for the Shapiro-Wilk test is below .05 (e.g., .032), it suggests that the data are not normally distributed. If you find that the variable representing the differences between the pre and post scores is not normally distributed, you could consider doing a non-parametric test instead (e.g., a Wilcoxon test).
@@DavidRobinsonPhD Hi, sorry to bother you again, but your advice was incredibly helpful last time! I have a question about my statistical analysis. For some of my secondary outcome variables, I used paired t-tests, while others required a Wilcoxon signed-rank test. My dissertation requires me to report confidence intervals alongside effect sizes. I understand that the formula for the effect size in the Wilcoxon signed-rank test is r = z/n, whereas, for the paired t-test, it's the t-statistic divided by the square root of the sample size. Can I use these two methods to calculate effect sizes in my write-up?😁
@@math6661 That's right, though just remember that you're dividing z by the square root of n to get r. Later versions of SPSS can calculate Cohen's d for you when you run a paired t-test: ua-cam.com/video/7RXnmQEBguU/v-deo.html
Thank you, this was really helpful
Thanks, glad it helped!
Hello,
I am writing my master dissertation at the moment, and your videos are of great help! Thank you very much :)
I do have a question though; I am sorry if the question is strange; I'm new on SPSS. I am doing both dependent and independent t-tests for my data, so I've watched your videos on each test. I noticed that you have conducted the normal distribution test differently in each video. Is it enough to conduct normal distribution test just one time, and then conduct paired t-test and independent t-test?
Hi, thanks for your question and glad to hear the videos have been helpful! It wouldn't be possible to check the normality assumption for both tests at once as the independent t-test assumes that the data are normally distributed within each group whereas the paired t-test assumes that the differences between the two time-points are normally distributed.
@@DavidRobinsonPhD This helps a lot!! I appreciate it, thank you sooo much!
Hi there! Your videos have been so helpful! I have a quick question I am currently analysing data for SD-WISE 7 item scale - I placed my pre and post-intervention scores into Shapiro-Wilk and It comes as normally disturbed (p = .032). Following this I would use a paired sample t-test? Just so confused haha any help would be appreciated :)
Hi, glad the videos have been helpful! If the p value for the Shapiro-Wilk test is below .05 (e.g., .032), it suggests that the data are not normally distributed. If you find that the variable representing the differences between the pre and post scores is not normally distributed, you could consider doing a non-parametric test instead (e.g., a Wilcoxon test).
@@DavidRobinsonPhD Honestly, thank you so much! Have a lovely week ahead of you!
@@math6661 Thanks, you too!
@@DavidRobinsonPhD Hi, sorry to bother you again, but your advice was incredibly helpful last time! I have a question about my statistical analysis. For some of my secondary outcome variables, I used paired t-tests, while others required a Wilcoxon signed-rank test. My dissertation requires me to report confidence intervals alongside effect sizes. I understand that the formula for the effect size in the Wilcoxon signed-rank test is r = z/n, whereas, for the paired t-test, it's the t-statistic divided by the square root of the sample size. Can I use these two methods to calculate effect sizes in my write-up?😁
@@math6661 That's right, though just remember that you're dividing z by the square root of n to get r. Later versions of SPSS can calculate Cohen's d for you when you run a paired t-test:
ua-cam.com/video/7RXnmQEBguU/v-deo.html