Forum Discussion
Preventing Invalid Response
Hello,
I have made a questionnaire that scores users on their hazardous attitudes based on likert scales. They are set up so that the user must score the most likely scenario as a "5", the next most likely scenario as a "4", the next most likely scenario as a "3", the next most likely scenario as a "2", and the least likely scenario as a "1". See the screenshot below. Each of the five possible scenarios (or answers) corresponds with a hazardous attitude, so if they select and submit the likert button 5 from the "anti-authority" scenario, that adds a value of five to the variable "anti-authority." If they rate the "invulnerability" scenario as a 3, that adds a value of 3 to the "invulnerability" variable.
Despite multiple written instruction slides and a tutorial slide, I am having a difficult time preventing users from either skipping rating an answer, or them scoring multiple answers as the same number. Because they're using a scale of 1-5, if they score the slides correctly, the variables should all be adding a total of 15 to the total score. (5+4+3+2+1). I've created a "total" variable that is the sum quantity of all the variables added together, and a way for the user to see an invalid response layer if they do not score their scenarios correctly. (i.e. making them all 5s, and making that total number be greater than 15.) However, if they go back to the slide and reset it, the damage has already been done to the total variable quantities, and they're just adding on to the mistaken total. Is there any way to reset variables to what they were on a previous slide?
RaeganMadoerin: That's an interesting challenge!
Using number variables and then totaling them isn't a good idea. After all, the total would equal 15 if I selected five 3's.
The attached file has a different way to verify that the user selects one button in each row and one button in each column.
The checking is done with triggers on invisible layers. One layer checks that each row has a selection. If not, it shows a Warning layer. If so, it "shows" the next invisible layer.
That invisible layer checks whether each column has a selection. If not, it shows the Warning layer. If so, it shows the Done layer. The Done layer is where NEXT is enabled. That layer's Slide Layer Properties prevent the user from clicking the base. So there's also a button to reset the important parts of the interaction if the user wants to change their answer.
I put programming notes under the slide area of those layers to explain what's happening.
The demo only has 3 rows and 3 columns. But the programming could be adjusted to account for more choices. That just means more triggers. 😁
6 Replies
My first thought was, "Wow, this is creative!" My second thought was, "Wow, this is challenging!"
You might be able to address the skipping problem by not "unlocking" the next item until the first item has been rated. One way to do that is to simply leave the items hidden on the screen until the immediate previous item has been selected and rated.
There are probably better solutions, but this might be worth exploring. Good luck!
RaeganMadoerin: That's an interesting challenge!
Using number variables and then totaling them isn't a good idea. After all, the total would equal 15 if I selected five 3's.
The attached file has a different way to verify that the user selects one button in each row and one button in each column.
The checking is done with triggers on invisible layers. One layer checks that each row has a selection. If not, it shows a Warning layer. If so, it "shows" the next invisible layer.
That invisible layer checks whether each column has a selection. If not, it shows the Warning layer. If so, it shows the Done layer. The Done layer is where NEXT is enabled. That layer's Slide Layer Properties prevent the user from clicking the base. So there's also a button to reset the important parts of the interaction if the user wants to change their answer.
I put programming notes under the slide area of those layers to explain what's happening.
The demo only has 3 rows and 3 columns. But the programming could be adjusted to account for more choices. That just means more triggers. 😁
This is GENIUS! Thank you so much, I'll try experimenting with this and get back to you.
You're welcome. Pay it forward when you can. 😊
Guess what...it completely worked! I did tweak one thing where for the row check layer, I just had it nearly identical to the column check layer... if none of the buttons in a row were clicked, it would show the warning layer. Thank you!
I'm glad you got your file working as desired!
Yes, the row-check layer could just check if any row had buttons that were all unselected. I used the value of the associated variables, since those were available.
FYI: I figured other folks might be interested in that type of interaction, and possibly even a graded version. So I updated the .story file and put it in a new post this morning: TIP: Ranking-Order Interaction | Articulate - Community
