Research data analysis in Excel: Column indexing - excel

I need to analyse empirical research data:
Given the following data:
The explanation: 5 people answered Question 1 with "entirely agree", which is equivalent to a score of 7. Zero people answered Question 1 with "entirely disagree", which would be a score of 0.
I want now to calculate the number of responses strictly less than the Median and the number of answers equal to the Median.
The result should look like this (the Median is given thanks to Research data analysis in Excel: Median of x times column value)
I tried with the functions =INDIRECT and =INDEX; without success.
Would be great, if someone could help me. Thank you.

Use SUMIF:
Below:
=SUMIF($B$6:$H$6,"<" & I2,B2:H2)
Equal
=SUMIF($B$6:$H$6,"=" & I2,B2:H2)

A working, but not a better solution is to sum up the row conditionally:
the highlighted cell has this formula:
=IF(B6<I2,B2,0)+IF(C6<I2,C2,0)+IF(D6<I2,D2,0)+IF(E6<I2,E2,0)+IF(F6<I2,F2,0)+IF(G6<I2,G2,0)+IF(H6<I2,H2,0)
I hope it helps.

Related

Using IF with ROUNDDOWN, To Calculate How Many Times For A Process

I apologize if the title is a bit vague. I am trying to create a calculator that takes into account how much "scrap" I have, how much is needed to resmelt it, and how many bars recieved.
Currently using:
(=if((amtOwned/qtyToSmelt)<1,,rounddown((amtOwned/qtyToSmelt)*barsMade))
Table and Formula
The problem I am having is you MUST have the QTY to Smelt. But the value returned includes partial quantities.
Ex. 125 Bottle Caps should equal 4 Bars total. Yet it returns 5.
How can i make the formula only account for increments of the bars recieved?
Thank you for any help, again i apologize if this isn't that clear. Im not exactly sure how to express my need in this situation.
I have tried messing around with the syntax and where every argument sits, even this formula is the most recent iteration of what i thought would be needed.
EDIT: I have tried using the TRUNC function and this seems to be working as I need it to. The formula now is:
=TRUNC((AMTowned/AMTneeded),0)*barsRecieved
=TRUNC((136/50),0)*2 This is returning 2 bars instead of 3. Which is exactly what I need.
It appears this is working by truncating the number first then multiplying it. So, 1.5 becomes 1 before being multiplied. This was my guess after doing more research. I had been searching for a while before I posted this but am glad to have learned what I have in searching for this.
There is a tool for auditing formulas. To see it go to Formulas > Evaluate Formula.
So here is you formula =IF((E3/C3)<1,,ROUNDDOWN((E3/C3)*D3,0))
Have you tried the calculation on your regular calculator? To me it is doing what you would expect. (125/50)*2 = 5

How to grade a test with different points by answer in Excel?

I'm trying to do something similar than this: https://exceljet.net/formula/score-quiz-answers-with-key
But it is not the same because I have different "points" for each question.
For example, correct answer for question 1 gives you 2 points, but correct answer for question 10 gives you 6 points.
I need to get the total of points for each student based on their answers and how many points each answer gives them.
This is the sheet with the answers and points:
Please try this formula. Paste to R5 (first result row) and copy down.
=SUMPRODUCT(--(B5:Q5=B$1:Q$1),B$3:Q$3)
It would probably be easiest to add more cells to calculate how many points each student got for each question. Then you can sum those up and divide by the total.
Here's an example.
screenshot
Give a try on below formula.
=SUM((B5:Q5=$B$1:$Q$1)*($B$2:$Q$2))
Matrix multiplication will also give you desired result.
=MMULT(--(B5:Q5=$B$1:$Q$1),TRANSPOSE($B$2:$Q$2))

Excluding 'N/A' answers in Excel table

Say I have a table of survey results. The scores for each answer are from 1-4 inclusive (1 worst, 4 best). An 'N/A' answer is represented as 5 - misleading when analysing!
I want to use a function in the final column that calculates the respondent's overall score for the survey. Rather than replacing 5 with 0 for a 'N/A' answer, I'd like to simply exclude that number altogether, and also exclude it from the overall maximum.
For example, say my maximum overall survey score was 80. If someone puts 'N/A' for one question, then I'd like their score to be n / 76, rather than n / 80.
Any suggestions? Feel free to ask for clarification if needed.
EDIT: See below for example image. I'm trying to get Overall Score / Max Score.
So after playing around with the available functions I stumbled across 2: COUNTIF and SUMIF.
As above, I'm trying to calculate Overall Score / Max Score.
So the formula I use is:
=SUMIF(range, "<5") / (COUNTIF(range, "<5")*4)
Stepping through the function:
=SUMIF(range, "<5") calculates the overall score (any number less than 5, ignores text)
(COUNTIF(range, "<5")*4) calculates the number of cells with numbers less than 5, ignores text. I then multiply this by the maximum score for a given question (which was 4). This gives me the maximum score possible given the number of 'N/A' responses.
Finally I divide the former by the latter and get the score I'm looking for. Hope this helps!

Removing Lower/Upper Fence of outliers from input data to then be evaluated

What I have attempted:
AVERAGEIF(B11:V11,">+MEDIAN(B11:V11)")
What I am trying to do:
I would like to take the average of the upper half of given data. Elaborating more. I would like to find a formula that will allow me to remove a given lower fence of outliers and dissect the data then given to me. I would greatly prefer to maintain this formula within one cell "not grabbing different results from formulas within multiple cells".
Update:
Following through I found the solution.. I think.
One thing I should have explained further:
The data coming in replicating a typical sqrt function.
What I wanted to achieve is to capture the mean of the "plateau" of the data.
The equation I used was:
=AVERAGEIF(B3:B62,(">"&+TRIMMEAN(B3:B62,0.8)),B3:B62)
This was something I just copied and pasted. of course "B3" and "B62" are significant only for my application.
My rough explanation of the equation:
TRIMMEAN will limit the AVERAGE to the top 20%(">")(0.8) of the data selected. So for my application, this SHOULD give me a rough mean of the "plateau" of the data i would like to find the mean for.
This formula calculates the Median() of the range, then AverageIf() uses the median and only grabs values that are greater than or equal to >= the median ~ giving you the average of the 'top-half' of your values.
AVERAGEIF(A1:A10,">="&MEDIAN(A1:A10))
Hope this help!

Set Pass/fail based on 10 answers from each of many individuals

I'm trying to output a Pass/Fail indicator for each student. A new student is on every row and their 10 answers are logged from Column E to N.
I want to match their answers with the 10 correct answers in Column S to AB, count them, and then have the word Pass show in the results Column O if they have a 70% pass rate, and fail if they are under that 70%.
All help appreciated.
I'd suggest a separate row for the answers but in matching columns:
then something like:
=IF(SUMPRODUCT(--(E$1:N$1=E3:N3))>6,"Pass","fail")
As #Jeeped points out, the correct answers could be left where they are with only a small adjustment to the formula above but (apart from the lurid colours!) I think the layout shown is neater and possibly more useful.

Resources