Statistical analysis on Excel

[Dalach]

Hi there,

I am doing a medical research project using excel based data, and was wondering if it is possible to perform simple stats on excel without using SPSS?

 

[SpillerBD]

There are numerous statistical functions in Excel.
Conrad Carlberg went so far as to wite a couple books on the subject too.
Statistical Analysis: Microsoft Excel 2013 | Que

 

[Dalach]

Though very thorough as I'm sure Conrad was in describing the subject, I was hoping to get some advice tailored to my project! I was told it would be reasonably staraightforward, but can't seem to make any stab at it myself as I am an excel novice.
Perhaps you might be able to help??

My project is on knee replacements and I want to do stats on three variables, which are SF12, WOMAC (outcome scores) and ROM (range of movement)
What i am doing is comparing pre and post op scores for each of these variables. (and doing the same for 4 different groups, to compare them across these variables)
I have all the data recorded but have no experience with exel so cant make head nor tail of it!

below is a copied and pasted sample of the data

[TABLE="class: cms_table, width: 742"]
<tbody>[TR]
[TD]Pre Op SF12[/TD]
[TD]Pre op WOMAC[/TD]
[TD][/TD]
[TD]6/12 post SF12[/TD]
[TD]6/12 post Womac[/TD]
[TD]preop ROM[/TD]
[TD]6/12 ROM[/TD]
[TD]ROM dif[/TD]
[/TR]
[TR]
[TD="align: right"]29[/TD]
[TD="align: right"]48.5[/TD]
[TD][/TD]
[TD]6/12 30[/TD]
[TD]6/12 96.1[/TD]
[TD]10 95[/TD]
[TD]5-120[/TD]
[TD="align: right"]30[/TD]
[/TR]
[TR]
[TD="align: right"]29[/TD]
[TD="align: right"]37.9[/TD]
[TD][/TD]
[TD]6/12 32[/TD]
[TD]6/12 87.1[/TD]
[TD]0-110[/TD]
[TD]0-130[/TD]
[TD="align: right"]20[/TD]
[/TR]
[TR]
[TD="align: right"]31[/TD]
[TD="align: right"]37.1[/TD]
[TD][/TD]
[TD]6/12 32[/TD]
[TD]6/12 90.9[/TD]
[TD]0-120[/TD]
[TD]0-120[/TD]
[TD="align: right"]0[/TD]
[/TR]
[TR]
[TD="align: right"]32[/TD]
[TD="align: right"]37.1[/TD]
[TD][/TD]
[TD]6/12 32[/TD]
[TD]6/12 62.9[/TD]
[TD]0-115[/TD]
[TD]0-120[/TD]
[TD="align: right"]5[/TD]
[/TR]
[TR]
[TD="align: right"]27[/TD]
[TD="align: right"]21.2[/TD]
[TD][/TD]
[TD]6/12 31[/TD]
[TD]6/12 90.3[/TD]
[TD]5 95[/TD]
[TD]0-125[/TD]
[TD="align: right"]35[/TD]
[/TR]
[TR]
[TD="align: right"]31[/TD]
[TD="align: right"]34.8[/TD]
[TD][/TD]
[TD]6/12 .33[/TD]
[TD]6/12 68.8[/TD]
[TD][/TD]
[TD][/TD]
[TD][/TD]
[/TR]
[TR]
[TD="align: right"]26[/TD]
[TD="align: right"]48.4[/TD]
[TD][/TD]
[TD]6/12 25[/TD]
[TD]6/12 91.4[/TD]
[TD]0-110[/TD]
[TD]0-120[/TD]
[TD="align: right"]10[/TD]
[/TR]
[TR]
[TD="align: right"]29[/TD]
[TD="align: right"]0.8[/TD]
[TD][/TD]
[TD]6/12 31[/TD]
[TD]6/12 97[/TD]
[TD][/TD]
[TD][/TD]
[TD][/TD]
[/TR]
[TR]
[TD="align: right"][/TD]
[TD="align: right"][/TD]
[TD][/TD]
[TD][/TD]
[TD][/TD]
[TD][/TD]
[TD][/TD]
[TD][/TD]
[/TR]
</tbody>[/TABLE]

What I want to do is find the average diference between the totals of two columns to start with. Any suggestions?????

Thanks in advance, Im happy to clarify if more info needed!

 

[Dalach]

Oh, forgot to add- I am only interested in those rows that have valus available for both pre and 6/12 post op columns.

 

[Dalach]

and i would be greatly indebted if you could take the time to help me as im very very stuck!!! thanks

 

[SpillerBD]

I would begin by making sure your data is in a table suitable for Pivot Table Use. No Blank Columns to begin with. Cells that should be read as zero should have a zero in it and not be blank. Blank cells are undesirable, but if there is not valid data for the value, then it becomes a different issue...
"average diference" should be a big deal. You can make a "helper" columns or calculated field in PivotTable. The helper field is easier to understand.

There are a number of "D" functions like DAVERAGE that you may be interested in...

 

[AhoyNC]

Another thing to look out would be paired t test in Excel.
Also, if you have access to SPSS I would use it to run boxplots of the pre & post data.
Try a google search on "paired t test in Excel".

Paired Sample t Test | Real Statistics Using Excel
https://www.youtube.com/watch?v=qClNnnigg6c
[url]https://www.youtube.com/watch?v=wy8GVt7Ityk

[/URL]

 

[Dalach]

Another thing to look out would be paired t test in Excel.
Also, if you have access to SPSS I would use it to run boxplots of the pre & post data.
Try a google search on "paired t test in Excel".

Paired Sample t Test | Real Statistics Using Excel
https://www.youtube.com/watch?v=qClNnnigg6c
https://www.youtube.com/watch?v=wy8GVt7Ityk

Hi!
Thanks for the advice, I had a look at those links but stil cant figure out what tests i need to perform, basically i have the data from my above comment for two cohorts and i want to compare the two groups across those parameters. So I think my null hypothesis would be that intervention x does not provide statistically significant improvement over intervention y.

But again, I have no idea what im doing so dont even know where to start!!!!!!

What tests do i need to do and why???
Response with any advice would be greatly apreciated, ive a deadline to have this writen up by friday!!

Thnka in advance

 

[AhoyNC]

See if any of these links help.
If you have data from 2 test for the same subject (pre and post) then I would look at paired tests.

Two of the links show how to make boxplots in SPSS and run a paired test in SPSS.

Boxplot in SPSS
http://geography.unt.edu/~wolverton/How to make a boxplot in SPSS.pdf
https://www.youtube.com/watch?v=gOj9wBv-IfM


http://www.statstutor.ac.uk/resources/uploaded/coventrypairedttests.pdf
Data Analysis - Paired t-tests
EXCEL Statistics tutorials paired t-test

Hypothesis Test: Paired Means\

 

[SpillerBD]

Statistics implies numerical analysis. The data you've presented has only three columns with numerical values.
The "ROM Dif" is a result of looking at a difference of values but not of regular numbers. Once you get to a good data set I think we can begin with some better examples of what can be done to yield your desired results.