Hi,

I am having problems setting up my survey to score a set of array questions and output a calculated score for the user. I have looked on the forum but cannot see how to set this up?

I have detailed the scoring/calculation below in stages.

Note: All questions are arrays with 7 options.
Stage 1: Set up each answer option so that it corresponds to a score. This set of scores is different for each subquestion.
Stage 2: Calculate total score from that question group (sum of sub-totals).
Stage 3: Set up an expression/formula where the scores from each question group are merged together to create another, final score. This final score will depend on which question groups the user has answered. This stage is I think similar to the 'concatenate' function in excel. The final score is the only output to the user.

Thanks for your help,
Jack

Hi, Jack,
I suppose you are a bit confused about assessment and expression manager.

1. You said "Set up each answer option so that it corresponds to a score. This set of scores is different for each subquestion."

How do you do that. In my opinion the answer options are always the same in an array question.


The best: Send an example of these questions in a small sample survey (*.lss)

Best regards
Joffm

P.S.
Just to add something:

This final score will depend on which question groups the user has answered.

Therefore there is the NAOK "No answer is ok"

Thanks for your reply.

"How do you do that. In my opinion the answer options are always the same in an array question." Perhaps each one of my subquestions needs to be a separate question in that case?

Survey is attached as (.lss).

Thanks for your help.

Hi, Question
and what are you going to do?
You said you want to get a score out of the array questions.
G1Q00001, G2Q00002, G3Q00002, G4Q00002, G5Q00002

Okay, first thing:
Your codes of your answer options are text "A1", "A2". How do you sum them?
So you have to change them to numerical values.


Second: G1Q00002 has a relevance equation "((Question1.NAOK == "A2"))"
What is "Question1"?

I attach a very rough example - sorry, no more time, nbut you should see, how it works.



Joffm

Hi,

Thanks so much for your help!

When I try to import this file it has error "Invalid survey ID" and wont load - are you able to change this your end so that i can see it?

I think i understand now that for each question to be scored differently as I want the questions must standalone and not be subquestions, even if they have the same answer options (which some do). I have made this edit on the 'Skin' (section 4) question group and now have the issue of duplicate answer codes! The scoring I want to implement has a coding system whereby multiple answers are assigned the same code.

eg Answer 1 = 0, A2 = 1, A3 = 2, A4 = 2, A5 = 2, A6 = 3, A7 = 4

Can you help with that?

Thanks again,

In this case you have to "rescore" your answer with an expression:

{Q1_SQ001Score=if(Q1_SQ001=="A1",0,if(Q1_SQ001=="A2",1,if(Q1_SQ001=="A3",2,if(Q1_SQ001=="A4",2,if(Q1_SQ001=="A5",2,if(Q1_SQ001=="A6",3,4))))))}

This you have to do for all subquestions. (And it might be different.)
Afterwards you can sum these new scores.
In this case you can have Skin20 and Skin21 in one array with two subquestions.


Now something general:

In your scale you have seven degrees from "not at all" to "a huge amount", or similar.
Do you expect people to be able to differentiate between "tiny", "small, "modest", "moderate"? IMO too much bias by personal language style.
And you seem to know that, because afterwards you reduce your scale to less options.

In this case I'd recommend a non linguistic scale from
0 = not at all
100 = ALL the time / a very huge amount, ...
this you can realize by a slider

Regards
Joffm

And please attach a *.lss file and not a *.txt file.

Thanks for you help with the rescores!

Thanks for your idea on the scale but this scale has been scientifically validated for a specific purpose and cannot be changed.

.lss file is attached, apologies.

Once I have mastered the rescores, how can I then sum these rescores for each question group and hide the answer from respondents but make the number available to use with further calculations? (Stage 3 in my original question).

Thanks