THANKS A LOT!!! I`ve been watching tutorials all week to get my results, this one is totally straight forward!! totally useful for a true beginner in statistics
I actually ended up taking the difference of the two scores and doing a Mann-Whitney U test, but I am going to check out your suggestion later. Thank you so much for the quick reply!
For paired T Test it is important to note that the data must be continuous in nature. Please note that in terms of some scores a person can obtain 65.5 but for some scores data is discrete ! Before applying T- Test please be sure of the data type and assumptions
thanks fr the respond:)....Yah i totally agree with you,BUT in order to confirm further i sent this video clip to a friend of mine and he as well agreed with you. However he just told me to confirm with you that when you REJECT the null hypothesis, does that mean you ACCEPT the alternative Hypothesis???
@MsReenaMann, you don't need to compare the t-value to a t-critical in SPSS. SPSS does it for you. The p-value (the significance value) is the number to look at.
I have a question… all of my participants did not complete the post-test. Can I still use the dependent t-test to test the null hypothesis? Thanks for the video!
We never really accept or retain or reject the Alternative Hypothesis, but if you reject the null, that means the evidence suggests the Alternative is supported. Thus, if you rejected the null hypothesis that men and women, on average, are the same height (and thus your alternative is that men and women are different heights on average), then that means you have evidence to support the Alternative Hypothesis.
thanks alot for the helpful video..But i am abit confused i attended an spss clss today and the lecturer was like if ur P-VALUE is greather than ur alpha (0.05) than u REJECT. Do u think he meant something else or??
So it is possible the lecturer misspoke or that you may have misheard him/her. In all cases, if your p-value is less than alpha (which in psychology is set at .05), then we reject the null hypothesis.
i am doing a problem with two sets of data ...a treatment added to the water to reduce plomb...number of observation not the same before and after...observations almost taken everymonth or every two months and there s note that at least one of the data set in sot normal? which test can be done?? it s t-test but not sure if paired or not
How to make sure that youry pre test and post test data are normal ? Would you like to help me about shapiro wilk normality test in that case please? Thank you...
Question: how could you include the effect of gender in the situation described in this video? Could gender be an interaction effect? Logically I would say no, genders do not change before or after the test. So if the samples are paired, then there should be no interaction effect of gender. Is this logically correct?
Ronnet You can, but not as a paired sample t-test. If you wanted to include the variable I described here but also include gender (male vs. female), you'd need to run a 2x2 mixed model ANOVA.
bernstmj Thx. But does it make sense to do so? Could gender cause an interaction efffectt on pre->post? Gender doesnt change between between moments so I wonder if there anything meaningful could be interrpret from such a test.
Ronnet So while gender doesn't change, it is possible that the effect that time has for men and women is different. Think about it this way. Maybe people just get better at the test from time 1 to time 2 (that would be a main effect), but maybe this is particularly true for women or only true for men. Maybe women get better but men get worse. While people's sex isn't changing, the way people perform on the test over time may interact with whether they are male or female.
bernstmj Thank you. I wasn't sure if that line of reasoning would work for dependent samples. It made perfect sense for independent samples but I couldnt wrap my mind around it for paired data. When you wrote it down the way you did, it suddenly made perfect sense. I guess I've been looking at my dataset for too long :P
I am doing pre and post study. I want post study to be more than pre. so should i give first post performance and then pre performance. Because t value shows post minus pre and if t value is negative it is not good.
If you have two variables (sex and time) then you need to run a mixed model factorial ANOVA. You will enter two variables, one for sex (male/female) and one for time (before/after). Watch my video on repeated measures ANOVA (ua-cam.com/video/WB3baLtBhTU/v-deo.html) and in the box where it says "fixed factors" put the "sex variable" there while you put "time" in repeated measures. I don't have a video for mixed model ANOVAs, sadly. Let me know if this helps.
Hi Bernstmj - You statistics are excellent. But I find this failure (in a broad scientific sense) that no one spells out the "Research Hypothesis". Typically, you wish to show that the RESEARCH HYPOTHESIS (the DESIRED RESULT) is proven. (This is always a one-tail test). But if you don't spell this out, very often the reason for the test - is lost, however good the statistics might be. (Just my rant about his "hole" in statistics).
THANKS A LOT!!!
I`ve been watching tutorials all week to get my results, this one is totally straight forward!!
totally useful for a true beginner in statistics
Short but delivers a punch. Very informative and direct and all under 5 mins. Thank you, kind sir 👌
I just LOVE, the way you explains things. You are a GREAT instructor!!!!!!!!!!!!!!!!!
I could watch this all day! A thrilling video with a compelling message.
so true
thanks for your help! ive been searching around for a good explaination for ages and this is the only one i understood!
good stuff
I rarely leave comments on videos but I just need to say THANK YOU!! YOURE A LIFESAVER
Thank you, this is the best resource online.
I could seriously listen to this man talk for the rest of my life. what a lucky wife he must have.
I really expected it to be more gravely but it's just nice
Excellent. Simple. Straightforward explanation. Thank you.
I actually ended up taking the difference of the two scores and doing a Mann-Whitney U test, but I am going to check out your suggestion later. Thank you so much for the quick reply!
YOU'RE MY VIRTUAL HEROE!!!!!
Huge Thanks! Life saver for me when I'm doing my Final Year Project.
this helped me study for my midterm! thanks so much!
Very crisp and easy description
Simple and straightforward. Thank you. I needed a refresher.
For paired T Test it is important to note that the data must be continuous in nature. Please note that in terms of some scores a person can obtain 65.5 but for some scores data is discrete ! Before applying T- Test please be sure of the data type and assumptions
Simplicity at its best!!!! Thank you...
Thank you, this video was super helpful, and I understood the concept. Thank you. Please continue to do more of these types of videos.
Life saver! thank you so much
thank you so much...you have saved me so much time and effort for my homework!!
Thank you so much. It helps a lot. More power to you!
Thanks buddy for the information 😊
Simple and concise. Great job. Thanks.
Love your videos! Thanks a lot!
Thank you very much. Your explanation was concise and on-point.
thanks fr the respond:)....Yah i totally agree with you,BUT in order to confirm further i sent this video clip to a friend of mine and he as well agreed with you. However he just told me to confirm with you that when you REJECT the null hypothesis, does that mean you ACCEPT the alternative Hypothesis???
@MsReenaMann, you don't need to compare the t-value to a t-critical in SPSS. SPSS does it for you. The p-value (the significance value) is the number to look at.
great video, how would you determine differences between gender (sex)?
Is there a need to compare T-Critical against the T-Value?
This is a great video, thank you for posting!
Thank you you saved my hours.
Thank you so much for this. If I want to do the paired t-test for only the male sex, please how do I go about it?
very helpful and clear thank you for posting this.
I have a question… all of my participants did not complete the post-test. Can I still use the dependent t-test to test the null hypothesis? Thanks for the video!
Great video! But is there a way to compare the before and after scores of different sexes is the same test?,
We never really accept or retain or reject the Alternative Hypothesis, but if you reject the null, that means the evidence suggests the Alternative is supported. Thus, if you rejected the null hypothesis that men and women, on average, are the same height (and thus your alternative is that men and women are different heights on average), then that means you have evidence to support the Alternative Hypothesis.
This is used for pre and post tests quasi experimental study?
THANK YOU LIFESAVER :D
thanks alot for the helpful video..But i am abit confused i attended an spss clss today and the lecturer was like if ur P-VALUE is greather than ur alpha (0.05) than u REJECT. Do u think he meant something else or??
Thank you, sir! God bless you!
this helped me a lot! thanks!
Thanks so much man!
very helpful thanks!
Can I know why the paired samples correlations table can be chopped off here?
THANK YOU SO MUCH SIR FOR YOUR VALUABLE SESSION..
The mean difference is negative. Can we reverse it?
So it is possible the lecturer misspoke or that you may have misheard him/her. In all cases, if your p-value is less than alpha (which in psychology is set at .05), then we reject the null hypothesis.
hi there, i would like to ask. if i would like to compare 3 years of pre-test agaisnt 3 years of post-test data, how should i do it?
Life saver!!!! Thanks :)
hey thanks a million fr the wonderful explaination:)
Can we use epidata v4.6 to create data templet in pre-post test expermental study design?
short and very helpful video
Would you please show how to get a representative graph out of this result that show the significance astriks on top and the error bars?
Thanks
thanks so much honestly!!
i am doing a problem with two sets of data ...a treatment added to the water to reduce plomb...number of observation not the same before and after...observations almost taken everymonth or every two months and there s note that at least one of the data set in sot normal? which test can be done?? it s t-test but not sure if paired or not
How to make sure that youry pre test and post test data are normal ? Would you like to help me about shapiro wilk normality test in that case please? Thank you...
How did you get the format so that it's not shaded?
Question: how could you include the effect of gender in the situation described in this video? Could gender be an interaction effect? Logically I would say no, genders do not change before or after the test. So if the samples are paired, then there should be no interaction effect of gender. Is this logically correct?
Ronnet You can, but not as a paired sample t-test. If you wanted to include the variable I described here but also include gender (male vs. female), you'd need to run a 2x2 mixed model ANOVA.
bernstmj Thx. But does it make sense to do so? Could gender cause an interaction efffectt on pre->post? Gender doesnt change between between moments so I wonder if there anything meaningful could be interrpret from such a test.
Ronnet So while gender doesn't change, it is possible that the effect that time has for men and women is different. Think about it this way. Maybe people just get better at the test from time 1 to time 2 (that would be a main effect), but maybe this is particularly true for women or only true for men. Maybe women get better but men get worse. While people's sex isn't changing, the way people perform on the test over time may interact with whether they are male or female.
bernstmj Thank you. I wasn't sure if that line of reasoning would work for dependent samples. It made perfect sense for independent samples but I couldnt wrap my mind around it for paired data. When you wrote it down the way you did, it suddenly made perfect sense. I guess I've been looking at my dataset for too long :P
Ronnet I totally get it. Happy to help!
T test means?
how can we control for variables on paired sample t-test?
how to compare the self resilience among orphan and non orphan children using spss??
How do I correct for serial correction using newey and west test when running paired sample t test for spss
Perfect.. thank you.
God bless your soul
so nice ... thank you
I am doing pre and post study. I want post study to be more than pre. so should i give first post performance and then pre performance. Because t value shows post minus pre and if t value is negative it is not good.
@Emilynicolewood, I am happy to help!
@cheyducky14, I am so happy they were helpful.
Thank you!
@tsp8s, I am so glad you like these. Please let me know if you have other questions or videos you'd like to see.
Thank you :D
Nice
... thank-you very much...
What if you wanted to see if there was a difference for men versus women
Ihad gone to India since 1990through F.O .Ihad analized my exprmental with T-test and Nova ,would receive certificate in statisc
Thanx
Thanks sir
If you have two variables (sex and time) then you need to run a mixed model factorial ANOVA. You will enter two variables, one for sex (male/female) and one for time (before/after).
Watch my video on repeated measures ANOVA (ua-cam.com/video/WB3baLtBhTU/v-deo.html) and in the box where it says "fixed factors" put the "sex variable" there while you put "time" in repeated measures. I don't have a video for mixed model ANOVAs, sadly.
Let me know if this helps.
Hi Bernstmj - You statistics are excellent. But I find this failure (in a broad scientific sense) that no one spells out the "Research Hypothesis". Typically, you wish to show that the RESEARCH HYPOTHESIS (the DESIRED RESULT) is proven. (This is always a one-tail test). But if you don't spell this out, very often the reason for the test - is lost, however good the statistics might be. (Just my rant about his "hole" in statistics).