My AS exam is in 3 days and your “A Level Statistics/Maths in Biology” playlist just saved my life. Your way of teaching is super calm and effective. I can’t be thankful enough!
amazing video, do we need to know how to do the calculations for the actual exam? I cant fully remember but i think i was told it was just to be used during the practicals
Hello, for the OCR examboard it is possible you could get a question on four different statistical tests. I am not sure about the other exam boards. Examiners will give you the formulas. It is probably unlikely you would have to do the whole calculation from start to finish (as in the video) because it would take too long but you may be asked to complete different stages of the calculation or they may ask which statistical test you should choose with a given set of data or even to interpret the results of an already calculated value. So you do need to know how to complete all the stages. Don't forget to check out the standard deviation video to learn a short cut on your calculator. Good luck 😊
Great explanations :) Helped me loads on finishing my required practical stats tests :) One question though - when doing the paired t-test, surely the n value should be n-1 instead? It's what the textbook says? (I could be wrong)
For a paired student's t test it is n-1 see video at 12:24. For unpaired, it is the sum of both samples minus the number of data sets as shown in the video at 6:25 don't forget there are two types of students t test we need to know for OCR. My video shows how to do both but maybe you didn't see my paired example at 12 minutes. Hope this makes sense 😊 so, none of us are wrong! Phew! 😁
Hi, im a little confused with the degrees of freedom for the students t-test, in my textbook it says df = (n1 + n2)- 2, whereas in your video it says df is n -1?
Hiya, my video covers the unpaired student's t test and then the paired student's t test. For the paired student's t test the df is n-1 which you have seen at 12 minutes 25 in the video. But earlier on in the same video I walk through the unpaired t test with a different investigation and for unpaired the degrees of freedom are n1+n2 - 2 See my video at 6minutes 13. So there are two different students t test in this video that need to be watched separately (don't fast forward 😂) otherwise you might end up with the start of one test with the end of a completely different one. Good to ask the question to clarify, I hope this helps 😊 good luck with your exams! 🍀
Thank you very much, i watched your video and it is really clear! Mrs H, i have a curiosity if you can answer to me. What happens if we have two samples of different sizes from two different populations (let's say 10 values from a population control and 3 values from a post treatment population). Does something change in the unpaired t-test formula to compute the critical value of the t? Thank you in advance 🙂
Hello! Because there are only 3 values in your second population I would suggest that would not be enough data to fulfil a statistical test. You would need more data. 😊
What if you were to say that there is a significant difference between the mean of the leaves? The issue there is the t-test value is to reject the null hypothesis?
Hi, yes, you reject the null hypothesis if there is a significant difference according to your calculated value being greater than the critical value 😊
Mrs, what's the difference between comparing 2 means using sd, and using t-test? I know you don't have to use the critical value for sd, since it's about drawing error bars I guess, but how would I know that I should the t-test instead of sd?
Hi, standard deviation helps you to know if the difference is significant or not if plotted as bars for example but when you use the t test you can see to what level the difference is significant. I think I am just saying what you have already said in your comment so I think you've worked it out by yourself. Great question that all students should be thinking about. Thanks for such a great question 😊
Unpaired 😊 will have to be the same exam though. There are probably more sophisticated statistical tests that would fit that kind of data but you will have to RU-vid a Maths teacher for that 😊
@@BiologyBreakdownwithMrsH Ok, Please lets consider this example… Below are the results for the same group of students before and after provision of free books. Clearly, results after provision of free books are much better but when I run the t-test (paired in this case), I’m told there is no significant difference between the two grades. What could be the cause? (G) (B) (A) A 2 6 B 3 7 C 1 18 D 8 6 E 12 1 F 8 1 G-Grade B-Before A-After
Hi, the students t test we do at a level biology compares two means, your data doesn't really lend itself to the t test because (if the numbers you are showing are means) you have more than two means. I guess you could do a chi squared. Compare your observed (after) with expected, so your null hypothesis being that there would be no difference between what you observed and what you expected, the expected for each grade would be a mean of the sum of the results - hard to explain like this sorry. But I don't think that will be right either, because you have before data too 🙈we only have to know a few statistical tests for Biology, I think there is a better one you could use but I don't know all of them. You will need to contact an A level Maths teacher who teach statistics in much more detail than we do. Phew-a tricky question for a Saturday night! 😊 I hope you manage to find your answer, sorry I can't help you with that particular data set. This is definitely for a Maths person 😊
Hi, you will need to learn how to do standard deviation on your calculator but the frustrating thing is that different models/makes of calculator work differently. My video on standard deviation walks through one model, a comment I’ve pinned is a different calculator so they might help you with what you’re looking for. What ever you do, remember to use the Sx at the end as in Biology we only ever work with a sample of data. Good luck!
@@BiologyBreakdownwithMrsH ohhhh i ended up figuring it out so after you find the differences of each data point you do difference - the mean ( x - x) ^2 and the sum of this for all the data points / 15- 1 is the SD ? Thats the correct method right ? so if i ever get a table where i have to differences in two pieces of data , that data becomes my 'x' in the standard deviation and i do the difference - the mean i see what you mean it would become tedious to work everything out
😊 yes, n-1, great work, being able to do it the long way is handy and definitely worth knowing how to do the quick way with your calculator in stat mode 😊
OCR A A Level, but I think other exam boards use this too! For OCR A, you could be asked a question from any point so we need to teach the whole statistical tests from start to finish.
Hi, when the value is negative you just ignore the negative sign. So say the critical value was 2.5 (I've just made that number up for the sake of explaining ) and you got a t test value of - (minus) 2.6. Ignore the minus/negative sign and treat that as 2.6, (only absolute numbers are relevant) this is greater than your critical value so you reject the null hypothesis. For A Level Biology that is as far as we need to understand. If you get a negative value in the spearman's rank correlation, you still ignore the negative sign when comparing to your critical value, if your spearman rank value is greater than your critical value then there is a significant correlation and if the value is negative you know that the correlation is a negative correlation. The chi squared test won't give you any negative values because of the process of squaring in the equation. These are the only statistical tests I am familiar with at the moment. It's quite tricky to try and explain without talking to you but I hope this helps a bit? 😊
@@BiologyBreakdownwithMrsH Thanks you for explaining. I was doing a t-test to find the significant difference between two means from different individuals.
I’m just confused how to put the numbers in stat mode because the frequency part is confusing what steps do I do on my Casio calculator (I have the older ones)
If you go to my standard deviation video description there is a method pinned there and a set of different instructions with a different calculator in the video. There are so many different models of calculator I don't know all of their unique workings. It might be best to check in with a friend who has the same calculator if neither of my methods work for you.
