Calculating T-Test Sample Sizes (with G*Power) | Statistics

[@peterdelaney2593]

Shared this with my lab group today after I covered the same material in about 4 minutes at the end of a long lab meeting. I think they may prefer yours.

[@DesigneLearningTutorials]

Let me know if I can create any more content for you! Subscribe to check out my latest series on usabiliy evaluation, including Stats made easy!

[@SamuelAlamo]

What a great video mate... So well explained, calmed, with examples... Best out there in Sample Sizes hands down. Love from spain!

[@DesigneLearningTutorials]

Much appreciated!

[@mazinalwish9506]

thank you so much

[@florence7091]

Really helpful video, thank you!

[@dr.raghadraad4403]

Hi, l'm study two groups first group drug abused second group health people and importance point that sample size for drug abuser is very limited in my country nearly 193 can you help my for calculated sample size and my objectives in my study 5 object how can calculated sample size for each objective ?

[@DesigneLearningTutorials]

Let’s solve this. Do you need to compare two groups for each of the five objectives?

[@izzatulyusuf5605]

Thank you so much for this! It was really helpful. However, Im still unclear with the ratio part, what if e.g im conducting a study to compare individuals from a single parent family and individuals from dual parents families, would this be considered as 2 different people?

[@DesigneLearningTutorials]

Yes. You have two different groups. One person can only be in one group. 100 People from a two parent family and 50 people from a single parent family is 2:1 ratio (0.5). I’m glad I helped you in your journey. Please like and subscribe to help other people find this video too!

[@izzatulyusuf5605]

@@DesigneLearningTutorials I see, thank you! So i would put in 2 as the allocation ratio? How do I determine which sample size is for which group tho. Like i would get 38 sample size for group 1 , and 76 sample size for group 2. Which one is which?

[@DesigneLearningTutorials]

@@izzatulyusuf5605 Hi, without knowing more about your data or what you have run it is hard to say, but it sounds like 76 + 38 = 114

[@Dani3pez]

This was extremely helpful! Thanks!

[@DesigneLearningTutorials]

You are extremely welcome Sir

[@polohou8919]

Very helpful, thank you

[@lolololeta1370]

Hi.i have study of effect of phototherapy on serum calcium in neonates.i have one group .and previous study measured serum calcium before +_sd .and serum calcium after phototherapy +_sd.how can i calculate sample size

[@DesigneLearningTutorials]

You need a paired samples t-test. Would you like a video on this?

[@snigdhodas2848]

what if i take the effect size based on a previous research and get an effect size as high as 2 or 3,should I transfer the same to the main window or take a lower effect size?

[@DesigneLearningTutorials]

Are you trying to calculate the sample size for a study based on the effect size of previous research?

[@snigdhodas2848]

@@DesigneLearningTutorials yes Sir

[@amsaran]

Can you please guide me if I want to use G-Power to calculate sample size (participant is drink escort)? I did find previous research which I could get around 3,200 samples have been done in the research. I need to compare two group of participants (escorts will be divided into two group for RCT). Could you please guide how to use the G-Power to calculate my sample size. Thank you in advance. Regards

[@DesigneLearningTutorials]

What test do you want to run? I imagine a t-test as in this video based on two groups?

[@jasonmiller6381]

Thank you so much for your tutorial. I learned a lot. I have a question for you, if t-test is the best or another option. Is it best to write here, or a diffrent communication channel?

[@DesigneLearningTutorials]

Hi Jason! I’m so happy to help you. If you have a question, just write it here. I may may a video answer, so please subscribe to stay up to date with all new content

[@jasonmiller6381]

@@DesigneLearningTutorials I have subcribed, your videos are very helpful. I have a study with 4 covariates as I was told pre-post test adds a lot of confounding variablbes. Do you have a video on IV and DV? - simplified? My study has 2 groups, 3 surverys, a prompt, then intervention and follow up with a prompt and 3 surveys. I am either doing a T-Test (Difference between two independent means (two groups) or F-Test (Ancova), unless you suggest a different test.

[@waynewu7763]

i have a different result when calculate power using norminv and normdist in Excel. I think its because i am assuming normal distribution. As I increase the sample size, I can see the distribution getting narrower. but i dont see that with GPower. I can see that GPower assumes T-distribution, but how is it determining the distribution without factoring standard deviation? Could you explain the math behind getting the beta value?

[@DesigneLearningTutorials]

Great question Wayne. Unfortunately I don’t know G*Power’s internal calculations, but maybe there developers can help you? www.psychologie.hhu.de/arbeitsgruppen/allgemeine-psychologie-und-arbeitspsychologie/gpower

[@waynewu7763]

@@DesigneLearningTutorials thank you!

[@DesigneLearningTutorials]

@@waynewu7763 I'm happy to have helped. Please like and subscribe to help others find this video

[@dr.raghadraad4403]

Hi, l'm study two groups first group drug abused second group health people and importance point that sample size for drug abuser is very limited in my country nearly 193 can you help my for calculated sample size and my objectives in my study 5 object how can calculated sample size for each objective ?

[@Daveaaaaa]

I didn't think you could make statistics easy to understand for someone who failed their Maths GCSE, but you managed it! :)

[@DesigneLearningTutorials]

Don't worry, I ran away from Maths A-Level to do Theatre!

[@musinzi55]

Or eskimos walked by.. 😂. I gotchu right away!!!

[@DesigneLearningTutorials]

My pleasure

[@mukeshmaharjan9449]

It was a really helpful video... thank you very much !!!

[@DesigneLearningTutorials]

Glad it was helpful!

[@6xn5]

Thank you.

[@gaynorayin5999]

Excellent video thank you so much!

[@DesigneLearningTutorials]

Glad you enjoyed it!

[@chaozhao6012]

Soooo helpful video. Looking forward to more stuff like this.

[@DesigneLearningTutorials]

More to come!

[@Chenxi-y1y]

Very helpful and easy to follow. Thank you ^_^

[@pervypirates]

Explained it so much better than any teacher I ever had, thank you

[@DesigneLearningTutorials]

Wow, thanks!