Hi, I’m new to surveygizmo and evaluating whether it can solve the following problem:
I want respondents to fill out a list, say a list of names of friends. I don’t know how many friends they’ll have, nor what their names are. Subsequently, I want to ask some questions about some of those friends depending on certain attributes.
Q1: List your friends. (Multiple text box question)
Q2: Which of your friends were at your birthday? (Checkbox question)
Q3: What did you get as a present from the friends who were at your birthday? (Multiple checkbox question)
The problem I seem to have, is that I can’t pipe answers to Multiple text box questions (or similar) like the names John, Bernard and Jack into the subsequent questions.
Can this be solved in surveygizmo and can you hint how I could do it?
Any help is greatly appreciated!
Hi Mart! That’s a great question, and I’d be happy to help answer. In short, piping from a textbox list is not possible because of how horrendous the data comes out.
A survey is only as good as the data you end up with. When looking at a regular radio button question in a CSV/Excel export, you end up with one column of data for all responses. A checkbox question will give you one column for each checkbox answer option, showing which options were selected. For example:
When piping, you are dynamically creating additional answer options. For a radio button question, that’s no problem. For a checkbox question, the number of columns grows in relation to the number of items you are piping in. If you were able to pipe from a textbox list, the number of items being piped in would be infinite (open text – respondents can fill in whatever they like). Instead of just adding rows to your response database with each new response, you would be adding columns as well:
This might look fine after entering one or two test responses, but after a hundred responses? Or a thousand? Just analyzing a single question would be a nightmare.
This is also compounded for each checkbox question you are piping to, leading to an exponential growth in the size of your response database — and don’t be too surprised if you find the limits of what our software can export, or a limit to the number of columns that spreadsheet software such as Excel can open!
In summary, lots of rows (responses) in your export are okay (linear growth), but lots of columns (exponential growth) are not. Or more simply, to have a happy surveying experience, always keep the end goal in mind
How to record this data safely
If you need to repeat a section of questions, don’t toss out piping altogether. The most ideal solution I’ve seen for recording data of this format is as follows:
(Q.1) How many ingredients do you have in your product?
> Number question type with validation (must be whole number, must be less than 10 or however many repeats of your data you are okay with ending up with)
> Use Page Repeat setting, based upon the value of (Q.1)
(Q.2) What is the name of ingredient number [page(“piped value”)]?
(Q.3) …questions about ingredient…
The data you end up with in your export will still be somewhat spread out, though much more manageable. Because you are piping off of a number, you end up with a finite number of repeats – no runaway columns!
Additionally, if you are on the Professional subscription, there is a Piped Data export that will allow you to end up with data from the example above in something like this format:
Here are a few resources in case you are unfamiliar with the features mentioned:
I hope this helps!
Thank you for taking the time to explain. I understand that open text piping may lead to a data mess. But in some cases, you can’t really get around it, because you can’t drill down the possible answers to a proper list.
The specific use case I have is that I have to ask producers about the ingredients of a product, and want to ask additional questions about each ingredient, depending on whether the ingredient is of animal origin or not. Basically, I’m trying to gather lots of rows in my export because I want to ask a set of questions for each ingrediënt, but because those ingredients are part of a product, I don’t want to tell my respondents to ‘fill out this questionnaire once for each ingredient in the product’.
But since there is no way of knowing beforehand how many and which ingredients are in the product, I can’t work from a checkbox or radio question onwards either. I could prevent the unlimited exponential growth of columns by limiting the amount of piped answers, e.g. by asking for a maximum of 5 ingredients, each to be filled in a separate single line text box. But if I understand you correctly, there is no way of piping an answer to an open text question into a next one?