Hello,
I'm fairly new to Dataverse and I've been researching enabling Make Data Count metrics in our Dataverse installation. I'm trying to fully understand how it would enrich our metrics, aside from adding in citations For the views and downloads, what exactly changes with those numbers?
I know, for example, that the classic counts do not weed out bots, but the MDC ones do? The MDC Dataverse documentations says that out of the box you get "Classic download counts", but what I'm trying to understand is the specific changes we'd see in those by enabling MDC.
Is a major part of the benefit of this is that DataCite can then show statistics as well?
I guess I'm trying to understand what this is really adding to our setup.
Thanks so much!
Bethany
Hi! Without MDC you don't get views or citations. And the download counts you get without MDC are pretty raw, shall we say. One of the benefits of MDC is that it tries to filter out double clicks, etc.
Thanks, Phil! This is helpful. Do the views come from the MDC code looking through the logs?
Yes, exactly. When the right setting is enabled, Dataverse starts doing extra logging.
Then those logs are analyzed by a separate application called Counter Processor.
Okay, makes sense. I know it's sending a SUSHI report to DataCite -- do those statistics show up anywhere on DataCite? I've been looking at papers from MDC enabled repos and not seeing anything relating to stats on DataCite Commons.
Yes! Absolutely! Let's take a look at https://commons.datacite.org/repositories
Here's QDR for example: https://commons.datacite.org/repositories/q98c6ry
127,330 views
Johns Hopkins: https://commons.datacite.org/repositories/g69cl6e
Nanyang Technological University: https://commons.datacite.org/repositories/4ygc74n
Thank you!!! That's one of the pieces I was missing. I was looking at the DOI for a dataset there, not the repository itself.
Thanks for helping to round out my knowledge. I had read all the Dataverse MDC docs and was missing a few key points (probably because I'm relatively new to the Dataverse scene).
Oh sure, there's a lot going on with MDC. There's even a new-ish tracker from DataCite ( https://github.com/datacite/datacite-tracker ) but it only tracks front end activity (like Google Analytics), rather than API activity, so our plan is to stick to the back end logging approach (sending those SUSHI reports).
Last question for now -- if we have mostly DOI's but do have some Handles, are the Handles just ignored? Or will the stats on pages with Handles just be off?
Handles are not supported by DataCite, sorry. Please see https://guides.dataverse.org/en/6.3/admin/make-data-count.html#limitations-for-dataverse-installations-using-handles-rather-than-dois
You could open an issue with DataCite to ask about Handles. :grinning:
:) Yeah, I saw they weren't supported. Do you know what that looks like when MDC is enabled and you have Handles in your repo? Are statisics on pages with Handles always blank? Just curious if you know if someone has done this before.
Hmm, at Harvard Dataverse we switched all our Handles to DOI, years ago.
Having multiple PID providers is quite new, only as of Dataverse 6.2, so I haven't heard of any stories of this yet. Or I forgot. :sweat_smile:
Actually, looking again at the chart at https://guides.dataverse.org/en/6.3/admin/make-data-count.html#limitations-for-dataverse-installations-using-handles-rather-than-dois I think the only think you'll miss is citations for datasets with Handles.
See how "MDC citations" is absent.
Yup - that would make sense.
Thanks for all your help with this! We'd like to enable MDC, but have a few things to figure out first. :)
Philip Durbin said:
Oh sure, there's a lot going on with MDC. There's even a new-ish tracker from DataCite ( https://github.com/datacite/datacite-tracker ) but it only tracks front end activity (like Google Analytics), rather than API activity, so our plan is to stick to the back end logging approach (sending those SUSHI reports).
Hello Phil, If our initial plan was to wait for the new DataCite tracker, perhaps we should consider moving forward with the standard backend log instead of waiting any longer ?
@GwenaΓ«l Doux sorry, I'm a bit confused, especially about who "we" is. :grinning:
Harvard Dataverse plans to process backend logs.
If you're saying your institution is also considering backend log processing, you'll be in line with Harvard Dataverse, at least the current thinking. Please see https://github.com/IQSS/dataverse.harvard.edu/issues/3
Sorry for the confusing pronouns :grimacing:
Ha, no problem. I will say that other GREI repositories are standardizing on that frontend tracker, last I heard. But again, we at Dataverse want to make sure downloads etc. from the backend (via API, for example) are tracked as well.
Actually, I just spoke with @Ceilyn Boyd and we might enable both. There are some details at https://github.com/IQSS/dataverse-pm/issues/196 if you're interested.
Last updated: Nov 01 2025 at 14:11 UTC