We recently launched a survey with unique links to a classroom of our participants. All used the same device types and all were accessing the survey at the same time. Some got messages saying "network error" and could not complete it; others were able to just fine. And one user who got the message refreshed his screen and was able to get in. So my question is this; has anyone else seen this sort of error, and have any idea what it is? When this happens, there's immediate panic and we start with the blame game; the people in the classroom swear it can't be the devices. The IT folks swear it isn't OUR network, it must be Qualtrics. App developers say it's not them, it must be the Wi-Fi. I care nothing about where the problem is, or about recriminations; I just want to figure out what's going on. So...anyone else experienced a "network error" message when using a survey, and know what caused it?
So I sometimes use the suggested topics tool in text analytics, mostly for when there is no pre-established library I can import. Sometimes they suggestions are good; sometimes, I just stare and wonder where the name came from. But that's okay, as we an rename them. But one thing I sometimes see is that two or more suggested topics have VERY similar keywords, and I'd like to combine them; not as a parent topic, but just a single topic. Is there anyway to just quickly combine two topics that way?
Working with a lot of EDI surveys these days. And I was wondering if anyone has seen or done a topic library on EDI that they've uploaded, and if so, would you be willing to share it?
I do consutling work and have a client who has the new XM directory; it's different than the old contact list setup in previous versions. Anyway, here's what's happening. In the past when I created a manual contact list for testing email distributions of a survey, I could do as many tests as I wanted. And each list would be stored as a contact list with ONLY what I put in it. What I'm seeing now is that if I make a contact list and put in a new embedded data variable (like a language code for testing), if I later put in a NEW MANUAL LIST, and do NOT put in a language code (or anything else), it will add the code used in a previous list using the same emails. So, if put in a test in spansih for doe.john@abc.com, and then do another manual list for the same email WITHOUT the language, it's putting the language in anyway. Has anyone else seen this or have an explnation for it?Thanks!
So math operations in Qualtrics are limiting to us, because they don't have an actual "average" function. Sure, you can take three variables and add them up and divide by three, but if someone doesn't answer a question, you get bad data.Here's an example. Say I want to average variables A, B and C. I'd do a math op that basically says X = (A+B+C)/3. And if A, B and C were all populated, it would work. But if not, Qualtrics doesn't treat the missing answer as a blank...it treats it as a zero, and throws your calculation off.Has anyone ever figured out a way, perhaps through custom code, to generate a true average from numerical data in a survey?
Okay. So I have a survey that uses a combination of a contact list trigger and authenticator to allow users to use one survey (set up) to populate the contact list with embedded data for another survey (end user). Pretty straightforward stuff. Using this solution during our move to virtual training, as it's much easier to deploy than emails.One of the questions we ask in the set up survey is the number of anticipated responses. That gets piped over into the embedded data in the contact list for the end user survey. In addition, a unique external data reference value is used to authenticate for that use of the survey. So, let's say one use of the survey expects 20 people to take it. That value and the external data reference they will use are popped into the contact list for the end users. What the users want is to know the response rate for each use of the end user survey. So, in the example above, say 15 people take it. That would be a response rate of 50%. But I would have
Okay. So in the research suite, there is a question in the standard Qualtrics library about country, with a list of around 200 countries in the world. In Vocalize, there is a map widget that includes a world view. But they don't exactly work together; for example, using that question, the map widget does not recognize the United States. I know you can use the three-digit country codes with that widget. But what I'm wondering (besides the obvious question of WHY these don't work together) is if the widget could be made to work off the recode values or variable naming for that question, rather than the displayed text. Then we could just put in the three digit codes, and that would fix most of it. Anyone tried that?
So I was doing some testing of a query string and a redirect to a second survey (long story, but it's how I can re-shape data for dashboarding). One thing I notice frequently is that if I delete a response from my data and analysis for my survey, it often continues to display in my dashboards. Anyone else ever experience that or have any suggestions for how to get rid of it?
So new reporting. You've seen it, you've tried it. But do you like it? Personally I find new reporting to be easier to teach to my novice users, but classic reporting is better for more complex or heavily customized reporting. There are elements to both I really like, but I'm curious what others think. Both can still be buggy (a new report I made yesterday steadfastly refuses to export to PDF), and I know that new reporting is being improved. But what features of classic reporting do you want to keep?
Already have an account? Login
Enter your username or e-mail address. We'll send you an e-mail with instructions to reset your password.
Sorry, we're still checking this file's contents to make sure it's safe to download. Please try again in a few minutes.
Sorry, our virus scanner detected that this file isn't safe to download.
