Biology paper 5 doubts

[Amy Bloom]

To find the degree of freedom you'll subtract 1 from the number of values you've got and I guess its aim is to reduce the error.
So for this question, you've got 20 values for each pH so this is how you get 38 : (20-1) + (20-1) and I guess if the question was to carry t-test on all 3 pHs you'd do that 3 times and degree of freedom will be a 57 instead
This is how I solved it, and I hope this helps

Sigh! thanks buddy, you have no idea how you relieved me from that doubt in me. I'm weak in stats so i didn't take s2 as part of my A2!

 

[xyz!]

The problem buddy is that i don't have any revision guide regarding s2 because i don't do s2. thanks at least 4 telling me that and good luck

ohh..ohk i'l tel u in short..
c error bars are basically sumthing lyk v do in phy..have u taken phy?
bt the only difference is that instead of doing +/- uncertainty in phy... for bio we +/- (2*standard error) ...and standard error is given by S/(root of n), where S=standard deviation and n=no. of samples...
hope dat helped!
gudluck!

 

[xyz!]

The problem buddy is that i don't have any revision guide regarding s2 because i don't do s2. thanks at least 4 telling me that and good luck

and btw i was toking abt biology revision guide..not s2! i myself havnt taken s2!

 

[xyz!]

can sum1 pls tel me the reasons for using chi-squared test and for using t-test for different situations? pls asap!

 

[Amy Bloom]

Hey for the same Oct/Nov 2010 P53 paper
look at question 1(c)(ii)
With which value should i compare with standard deviation in order to say that the std dev is larger or smaller hence results reliable or not??

 

[Amy Bloom]

and btw i was toking abt biology revision guide..not s2! i myself havnt taken s2!

Okay i don't have any bio revision guide. i'm doing all papers on my own actually.

 

[Amy Bloom]

can sum1 pls tel me the reasons for using chi-squared test and for using t-test for different situations? pls asap!

chi squared test is for discrete data.
t-test is for continuous data.

 

[Amy Bloom]

For paper Oct/Nov 2010 p53
Number 1 (c) (iii) do error bars involve ONLY standard error???

 

[xyz!]

Hey for the same Oct/Nov 2010 P53 paper
look at question 1(c)(ii)
With which value should i compare with standard deviation in order to say that the std dev is larger or smaller hence results reliable or not??

dat question is not related to the data..dey hav asked a general question..so the std deviation shows the spread of data from the mean!

 

[Amy Bloom]

dat question is not related to the data..dey hav asked a general question..so the std deviation shows the spread of data from the mean!

Yeah i know that but how can i qualify that the standard dev is small or big??

 

[xyz!]

Okay i don't have any bio revision guide. i'm doing all papers on my own actually.

oh..ohk..i hav posted wt is std error..pls hav a luk..
and yeah for dat questn in 53/o/n/10 we only got to +/- (2*std error) for the error bars

 

[xyz!]

Yeah i know that but how can i qualify that the standard dev is small or big??

u cant actually qualify it...
its lyk..ven u r given 2 different data sets...u got to compare der std deviations..the one wid a greater value of std deviation out of the 2..wil b less reliable..

 

[Amy Bloom]

oh..ohk..i hav posted wt is std error..pls hav a luk..
and yeah for dat questn in 53/o/n/10 we only got to +/- (2*std error) for the error bars

Okay then +/-(2*std error) is applicable in all cases then. You can also simply say that +/-(standard error) also, right? please confirm.

 

[Amy Bloom]

u cant actually qualify it...
its lyk..ven u r given 2 different data sets...u got to compare der std deviations..the one wid a greater value of std deviation out of the 2..wil b less reliable..

You mean to say that it depends on the data given to me, there is no generalised concept for this.

 

[Kumkum]

http://www.xtremepapers.com/papers/...and AS Level/Biology (9700)/9700_w08_qp_5.pdf
can someone please help me with question 2 (b)(iv)
which probability of "t" do i use to calculate confidence interval and btw the degree of freedom is 19..

 

[xyz!]

Okay then +/-(2*std error) is applicable in all cases then. You can also simply say that +/-(standard error) also, right? please confirm.

no as far as i hav read it is always +/-(2*std error) and not just +/-(std error)
and yeah...it depends on the data given to u...no generalized concept for it as u asked!

 

[Amy Bloom]

ohh..ohk i'l tel u in short..
c error bars are basically sumthing lyk v do in phy..have u taken phy?
bt the only difference is that instead of doing +/- uncertainty in phy... for bio we +/- (2*standard error) ...and standard error is given by S/(root of n), where S=standard deviation and n=no. of samples...
hope dat helped!
gudluck!

Whilst scrolling, its now i saw ur post.
Thanks. no i don't do phy.
it indeed helped.
+/-(2*std dev.) is always the case for error bars then?

 

[Amy Bloom]

no as far as i hav read it is always +/-(2*std error) and not just +/-(std error)
and yeah...it depends on the data given to u...no generalized concept for it as u asked!

Sigh! thank you ^^

 

[Amy Bloom]

http://www.xtremepapers.com/papers/CIE/Cambridge International A and AS Level/Biology (9700)/9700_w08_qp_5.pdf
can someone please help me with question 2 (b)(iv)
which probability of "t" do i use to calculate confidence interval and btw the degree of freedom is 19..

You use the t-critical value at degree of freedom 19, i.e, 2.09

 

[Kumkum]

You use the t-critical value at degree of freedom 19, i.e, 2.09

when choosing the probabilities do you always use 0.05 or does it depend on the confidence interval?