The Paired-Difference t Procedure

I discuss an example of the paired-difference t procedure, working through a confidence interval and hypothesis test and interpreting the results. (Paired-difference procedures are sometimes referred to as matched-pairs procedures, depending on the setting.)

The identical twin-schizophrenia data in this video is from:

Suddath, R., Christison, G., Torrey, E., Casanova, M., and Weinberger, D. (1990). Anatomical abnormalities in the brains of monozygotic twins discordant for schizophrenia. New England Journal of Medicine, 322:789--794. Values used in this video are simulated values based on the summary statistics found in the paper.

[samad.chouihat]

the best statistics channnel ever . thnks for the help. greeting from algeria

[Basenbloggar]

Wonderful! My professor forgot to mention that paired difference is useful when the parameters are independent. Homework suddenly turned very easy now, cheers mate!

[piarodriguez]

hi. I just wanted to say that your videos are really good. Im failing stats for sciences at the moment. We are using the Navidi book for Engineers and scientists which is pretty much Chinese mandarin for me, but your videos are helping out a lot. Things make sense now. Thanks!

[jbstatistics]

I got that using software (R), but you could also find that the value is close to 0.01 using a t table.

[hikariwuff]

I'd like to reject my null hypothesis that I will pass this class

[SirGamestop]

Good vid dude. This beats reading the textbook

[jacquelineoliver]

your videos are amazing its helped me so much with my stats class thank you

[Pheshen]

If I measure every pair not once, say 7 times across the week. How then does the calculation work? Do I need to average the 7 numbers of each pair and calculate the usual way?

[virtuousjoffrey]

My problem is identifying between dependent and independent populations and hence samples, can you help me out in identifying them without any problems?

[stxroshan]

If sample size is large, do we still use the t test statistic? Or do we change over to the z test statistic?

[andrescheepers]

Hi JB, love your videos! I didnt quite catch your last point about how this sample isn't random and thus doesn't quite generalize to the larger population. Do you mean that a truly random sample would have sampled sets of twins randomly, capturing both normal and half-schizophrenic sets of twins?

[kodido]

Thank you very much, very clear explanation, much appreciated !

[niki-mendozacatalan]

So do we fail to reject the null hypothesis?

[johnxavier]

Wow Understood in just a minute !! Thanks :)

[winterfell]

Thank you! Just one question, how did you get the p-value of 0.01? Thanks :D

[brentwalker]

It seems like he is developing schizophrenia as the video progresses, nice helpful video though.

[harmannd]

how did you get the p value to be 0.01? when i look on the z score chart 3.0= 9.987 (1-9.987)=0.001 isnt that the p value?

[Xerathful]

Don't we have to do affected-not affected when we calculate the difference?

[diana]

Why don't you use the formula with sigma to find standard deviations of differences

[anhtrinh]

how did you find the standard deviation?