General Betting
There is currently 1 person viewing this thread.
01 Aug 11
12:00
Joined:
Date Joined:
13 Apr 04
Apologies that I posted this on chit chat also but I feel that there may be a few more people here who could help:
Hello all, I wondered if someone would be able to help me with a piece of coursework I have to do, (hopefully) I won't bore you too much with the details but here is a little background:
Peak expression of a protein is shown at the same time every day, the study investigates the effect of two neurotransmitters upon the peak time of expression of this protein. The study looks at the effect of neurotransmitter A, neurotransmitter B and also the effect of Neurotransmitter A and B co-administered. The results are shown as time in minutes, a positive value indicates an advance in peak expression of the protein and a negative value indicates a delay. This study is carried out by administering the neurotransmitters at two different times of the day (early morning and late at night). My task is to correctly transcribe the data into a suitable package, display it graphically as best possible and then statistically evaluate it using the correct/most relevant tests.
Any pointers? Thanks very much in advance
Hello all, I wondered if someone would be able to help me with a piece of coursework I have to do, (hopefully) I won't bore you too much with the details but here is a little background:
Peak expression of a protein is shown at the same time every day, the study investigates the effect of two neurotransmitters upon the peak time of expression of this protein. The study looks at the effect of neurotransmitter A, neurotransmitter B and also the effect of Neurotransmitter A and B co-administered. The results are shown as time in minutes, a positive value indicates an advance in peak expression of the protein and a negative value indicates a delay. This study is carried out by administering the neurotransmitters at two different times of the day (early morning and late at night). My task is to correctly transcribe the data into a suitable package, display it graphically as best possible and then statistically evaluate it using the correct/most relevant tests.
Any pointers? Thanks very much in advance
sort by:
Show
per page
Replies: 25
|
By:
The study looks at the effect of neurotransmitter A, neurotransmitter B and also the effect of Neurotransmitter A and B co-administered
Guess first off you would want to know if there is any effect. ie a null hypothesis |
|
By:
i think you might need to "bore us with the details" should you want a proper answer.
when you're talking about being shown the peak expression time, is it a distribution or a discrete value? |
|
By:
Apologies DStyle, the reason I ommitted most of the details was in order to not deter people who may be able to help, I felt I had provided enough information, will try to fill in the blanks;
What I (incorrectly  )wrote as 'peak expression time' actually refers to the time of peak expression, it is a time value and the result values for each neurtotransmitter are expressed in minutes from this time, here is an example of some results.Early Morning All values in minutes A A+B B -94.2 -19.4 -19.2 -103.2 -22.2 -18 -45 -7.4 -3.2 -61 -20 39.2 |
|
By:
but is the peak expression time a distribution or a discrete value?
|
|
By:
It is only a small piece so the only information I am given is that "following stimulation the peak could appear earlier or later, the degree of advancing or dalaying can be calculated by subtracting the time of the peak before the treatment with the time of the peak post treatment, a positive value indicates an advance and a negative value indicates a delay in expression times' That is all I am told
|
|
By:
ok. so we're saying that the peak expression time is a discrete value which shows no variance.
how much data have you got? |
|
By:
Not a lot, I have 10 rows of date for morning and 10 rows of data for evening. I am not sure what is the best way to go about this, I was looking forward to a statistics module but I don't enjoy this
 |
|
By:
Each row has a time value for A, B, and then finally A + B
|
|
By:
can i ask what you've covered in your course so far?
|
|
By:
Some background stuff and then a look at:
z-test one sample t test paired t test independent sample t-test ANOVA chi-squared |
|
By:
as i say the problem here is the rather filthy assumption that the peak expression time is a discrete value. it almost certainly isn't. it is far more likely to be modelled by a distribution.
your assumptions are that your results with the NTs are normally distributed, and as you don't know the variance and you've got lots of comparisons it looks like an ANOVA. it'll tell you whether there's a different effect amongst the three conditions. |
|
By:
you can also do t tests for each group against the mean as well.
|
|
By:
Thanks for your help DStyle, would a t-test be looking at each group and then determining whether the NT did have a statistically significant effect?
I'll have a look at an ANOVA also, I may come back to you, apologies for being a nuisance, I didn't study pharmacy for my love of statistical tests but funnily enough if it's betting related I'm a little more interested |
|
By:
the reason they haven't given you the distribution of the normal peak expression time (despite the fact that there will be one) is that they want you to use one of the t-test and/or ANOVA.
stricly speaking, by giving you six different groups to compare i supsect they want you to use ANOVA. |
|
By:
And the use of ANOVA is just done using a statistical program or even Excel, is that correct?
|
|
By:
that sentence doesn't mean anything.
i'd suggest you go back to your lecture notes on ANOVA and work out what you need to do. |
|
By:
the best answer is always you dont have enough data.
|
|
By:
ANOVA sounds like some virulent disease in a sci-fi novel.
It's humbling sometimes to read this type of thread just to realise how many things there are out there that you know absolutely nothing about. |
|
By:
not sure when Analysis of Variance should be utilised on such a small data-set.
by definition, it defeats the purpose of obtaining an accurate measure of statistical significance. would explore chi-squared myself. for an example of same, look at CX Wong's book on the Statistical & Mathematical Handicapping of Horse Racing ["Precision"]. [chapter relates to the strike-rate of a Honk-Kong jockey at a particular track; results compared vs. p-value.] HTH |
|
By:
The Formula for Success is s = success, y = data, t = time
s=y*x*t, and x= "you exponentially never have enough" |
|
By:
bring in the main frame
|
|
By:
"formula","analysis method" and "data type" are all useful but arriving at the same answer as everybody else may not always be "the best solution"[;)]
|
|
By:
I understand statistical analysis but I love philosophy
|
|
By:
steeringjobnap has the best answer
|
|
By:
Careful you might give away a smidgen too much.
|