Millisecond Forums

Okay - I've done all that. Let's see what happens during the Friday pilot. I'll let you know whether we encounter any irregularities.
Thanks again for the quick support.
Marije

> Both, me and my colleague who used different machines, were likely to use it while there was a vpn connection to our university. Is that something to take into account?

Possibly, it would depend on how exactly that network is set up and how it handles caching.

If the file is cached locally, you can clear the systems' temporary files -- either via standard tools provided by the operating system, or (perhaps more conveniently) by running an app like CCleaner (download the "portable" or "slim version https://www.piriform.com/ccleaner/builds ; it's also available for Macs https://www.piriform.com/ccleaner/download?mac ).

> or am I missing something?

No! I missed something:

<openended f_practice>
/ recorddata = false
...
</openended>

The setting in the <openended> element overrides the setting at the <block>-level (as it should). This also needs to be set to true.

Thanks! Can I do anything to have the cache emptied here? Both, me and my colleague who used different machines, were likely to use it while there was a vpn connection to our university. Is that something to take into account?
Btw. I added the 'record data.true' to the practice block and ran the digit span, but I cannot find the recorded data in the raw data file - or am I missing something?
Marije

Thanks for the additional details. 4 out of 10 times seems high, so I'm wondering why I'm not seeing the problem. I'll keep digging and let you know what I find (if anything).

On the off-chance that one of the script's previous revisions that were uploaded at some point contained a mistake in the relevant section, there is at least a theoretical possibility that the "wrong" file is still cached somewhere and occasionally executed -- it could be cached either locally in the two systems' temp directories, or it could be cached and served by a caching server on your network. I am reasonably confident that the file is not cached and potentially still served by any of our systems.

I have removed the link to the launch page from the thread.

Hi Dave,
see here my answers:
> it did not happen always, but just occasionally.

Can you give me an approximate figure, i.e. 1 out of X times?

It happend, I think, 4 times in about 10 trials.
Finally,
(1) Are you seeing this on one specific computer or across multiple machines?
We have been using two different machines: Mac and a Dell with Windows 10.

(2) Was there ever a different version of digitspan.iqx uploaded that may have had some mistake?
Yes, I have changed it several times over the course of the last two days, but I have not changed anything to that particular section in the script

(3) Would it be possible for you to set
I've changed the practice block record data to 'true'. We're doing a life pilot on Friday, so that will give us plenty of data.
I'll let you know whether it keeps happening.

Btw. is there a way to delete the post with the link to the web startingfile? I'd prefer it not to be accessible for outsiders.
Thank you again for the quick support. This is also why we pay a license fee - it is worth every penny!
All the best,
Marije

> Btw. it did not happen always, but just occasionally. That's what puzzled me, because if it were in the script, it should happen always, shouldn't it.

Yes, that's exactly what makes it super weird. I've been up and down through the script multiple times now, and there literally is no code-path that could cause <openended f_practice> to display 4 digits or cause <block f_practiceblock> to somehow execute any other trial than a single instance of <openended f_practice>.

> it did not happen always, but just occasionally.

Can you give me an approximate figure, i.e. 1 out of X times?

Finally,
(1) Are you seeing this on one specific computer or across multiple machines?
(2) Was there ever a different version of digitspan.iqx uploaded that may have had some mistake?
(3) Would it be possible for you to set

<block f_practiceblock>
/ preinstructions = (f_practiceinstructions)
/ trials = [1 = f_practice]
/ recorddata = true
/ postinstructions = (practiceend)
</block>

in the forward practice block so that its data will be captured in the raw data files? That might provide a lead, and that data is not captured currently.

Thanks!

Hi Dave, I really appreciate your quick support. See her the starting page

As said, instructions are in Hungarian. You need to enter a 5 digit code TWICE to start it - please use any ID starting with 7.
Thanks!
Btw. it did not happen always, but just occasionally. That's what puzzled me, because if it were in the script, it should happen always, shouldn't it.
I've added the file again -as you see modified today at 1.30 local time.
Best wishes,
Marije

Can you direct me to the web experiment's launch page? Also, have you -- at some point after initially uploading the digit span script -- replaced the script with a revised version? Thanks!

Hi Dave, yes I am. I just double checked. As this is the practice trial, do you think I might just increase the charlimit to 4? At least the practice wont give people wrong feedback then. What do you think?

I'm not seeing an obvious bug in the script and despite running the script a few dozen times I've never seen 4 digits displayed in the forward digit span practice trial -- always 3 digits. Are you 100% certain the script you attached is identical to the one you have deployed online?