How can you compute the difference X naught reference point?

I allow setting process to happen only for two days or if I see the block of the seven days so one incidence is two days say a second is the same block I see for the strength seven days down the length I take a block certain 50 blocks I have pouring done setting then after two days I see the strength after seven days for the same doctor I see the strength okay again this is a before-after after is after five days and before this on the second day can I do that so again this what is the objective of this kind of a test the objective of this kind of test is to see whether we should allow it to sit for seven days of whether we should allow it to set only for two days.

If it happens in two days my productivity will increase right but if I were to give seven days which means I require more space to keep them right so if my work I get the same strength in two days as compared to seven days I would have done that only in two days but if I don't get the same strength after two days as compared to seven and I have to let it settle for seven days as simple as that so again before after right so these are applications so here I have taken a simple example because this we can very easily relate my scores before of the may wait before going to the gym after going to the gym my medical reports before taking the medicines for my ailment after taking the medicines, okay all these are applications.

I'm going to read as it's a students go where they are to column before and after if we do a description descriptive statistics we see mean is coming 18 a score after the training is coming 20 not much difference but is this difference statistically significant is the cushion visibly I feel eighteen point two and twenty point five not much right but is it statistically significant visually descriptive stats how do you present this information using a boxplot okay so syntax is given boxplot syntax is also given check for normality once again t-test applies when data is normally distributed so Shapiro test it does not seem to be a perfect normal but still I'll assume it normal because the p-value is saying it is normal okay if I would have gone with a histogram probably I would have concluded it is not normal right but the p-value is saying it is normal so let's assume normal this is a real relative test t-test one sample t-test independent t-test relative no because you are seeing the same data relative.

So t-test Arielle I give after before the two information is a two-sample outcome and the p-value is coming as point zero two which anyways from the score was much not much different but here I get a p-value which is point zero two so what do I conclude is there a significant difference in score before after based on statistical test or there is not much significant distance in score p-value is point zero you know alpha is point zero five we reject the null I accept the alternate and reject null means in t-test knowledge PA equal to PB mean equal to mean be alternatives not so there is a significant difference so that's our conclusion how we could have done the same calculation different way the data is of the same samples details of the same sample so what I can do is first compute the difference the moment.

I compute the difference mean becomes zero if there is no difference from a mean should be zero that to sample paired sample become one-sample t-test are you getting the point that's what I explained when I gave the formula, yeah M is the mean and s is the standard deviation of the difference T the moment you compute this becomes the same formula in this formula X naught is a reference point now I'm going to compute the difference X naught reference point becomes zero weight X naught reference point becomes zero so this formula writing as M minus X naught X naught is zero this formula and this formula is same so a paired t-test can be converted into a one-sample t-test and that's what I have shown how I can do the same calculation I do a difference between the despair and then apply one sample t-test I get the same p-value.

Your difference point is zero because you are null hypothesis assumption is sample a equal to sample B that is done the hypothesis sample H naught is sample a sample a mean equal to sample B mean sample B is here a before and a after in this case right do you agree this moment I moved to left hand side it becomes sample a before - a after equal to 0 H I H naught becomes so now if you want now if you want that difference will be more than two levels there will be more the difference will be you assume go to Hawaii to a score increase of two will be there if a score increase of two is there I assume it is no difference but marginal improvement will happen as I said in this frankly speaking if 18 point when 20 point why there is no use of the training read only marginal difference has to go what I could have done when I compare convert this when I convert this into one sample what I could have done is I would have said score a after before - - I add one more element - - now still mean is zero because I have already said minimum increment of two points will happen for me if the improvement is more than two only then it is a significant implement only they're not training is significant.

So I add that - two-factor automatically again mean goes down to zero no you are not taking with respect to - does that answer your question see that's what I am answering the objective is see you go to a gym as I was saying that hey you go to a gym you follow a diet plan you follow the exercise regime automatically will lose weight right you don't follow you're not going to lose right now what which factor is another control of the gym the exercise because if you come to exercise regularly we will know you are doing exercise but the diet plan whether you follow or not I don't know okay now but I want to take benefit of both the elements but one element is not in control so instead of saying I will because of my jimang and I plan 5 kg is what is lost so my target is 5 kg ok so I can do that calculation before after and - it with 5 kg the mean will now come down to zero so I can take that as a reference so that adjustment what you are asking you can do in this formula straightaway you take the digestion factor here and make it as one sample done right or wrong yeah so that is this test is specifically designed to say there is a difference between the two numbers.

But if you are not happy that two numbers are definitely different or not you want that hey it should be at least this much is the bare minimum I am expecting if it is about that point also then I will say it is you put that factor adjustment factor and make it one sample T - go ahead and do the test yeah so if the distribution is not normal if the dish Oshin the t-test makes use of make use of the fact that the distribution is normal however if the distribution is not normally a variant of the D test which is a nonparametric test which is called Wilcoxon details I am NOT a highlighted parametric and non-parametric so now I'll just highlight the term because if I would have highlighted this parametric and non-parametric term very beginning of the accession we would have got confused tests are classified into two one is called a parametric test and second is nonparametric test a test which makes use of mean.