0

We're looking to overhaul our filtering for our users. As a starter we were looking to do some initial research online. I've hit a bit of an issue of understanding in doing so.

We currently operate 'subtractive' filtering:

  • By default, everything is checked, user unchecks the data points they don't want to see.

Just a wireframe

Contrast to 'additive'/traditional filtering:

  • default everything is unchecked, nothing is filtered in the group you've filtered unless something is applied. (below the filtering can be found on John Lewis' website)

enter image description here

I can't really find any information online, or uses cases as to when you'd use the first model, ideally we'd have it documented as to why this decision was made, and if not find out from our users, but that wasn't the task I was assigned in this overhaul.

The only scenario I could think of is if your users want to only filter out 1-2 areas of data, most of the time. Rather than filtering into specific areas, but this is just a scenario I've made up. Does it also just go against any universal standards of filtering?

Any help would be greatly appreciated.

4
  • 1
    I'm guessing you never see what you call "subtractive filtering" because filter by ALL is actually no filter at all. Commented Nov 10, 2022 at 21:21
  • You say it's a guess, but if "users [actually] want to only filter out 1-2 areas of data" then provide an EXCLUDE list. Commented Nov 10, 2022 at 21:22
  • Yeah I wondered about this, is there an obvious pattern i'm missing such as an exclude list, I observed a few users remotely and they frequently unchecked all and re-ticked the boxes they wanted to see... Kind of the point of filtering right?! Commented Nov 11, 2022 at 10:41
  • "I observed a few users remotely and they frequently unchecked all and re-ticked the boxes they wanted to see..." There you go. I expect that's typical. The "Exclude" option could be provided in an "Advanced Filter" option that is closed by default—IF user research demonstrates a need. Commented Nov 11, 2022 at 15:25

3 Answers 3

1

As I did some research on this topic a while ago, checking different eCommerce sites, apps and talking to people - I would say: It does not really matter if you are going additive or substractive.

My outcome as an example:

enter image description here

0

I think you are mixing two different elements, the already-seen option's tags (source/sentiment) and a filter (time/location).

I would separate the filter from the tags but in the presence of all exposed elements.

In this example, all sources/tags are selected except Telegram (or vice versa, depending on the design):

enter image description here

0

Consider the user's need, what is the use case? Do users need to quickly exclude one or two items or to narrow the search to one or two items? Your first mockup might have users simply want to exclude some sources, to reduce the noise some specific sources may produce in results. In your second mockup, users might want to select specific brands they want to see results from.

Selecting ALL may automatically deselect or select all other options, depending on your user needs: enter image description here

enter image description here

1
  • "Selecting ALL may automatically deselect or select all other options, depending on your user needs: " — I don't understand your suggestion. Please elaborate. What exactly is clicked in each graphic? The "All options" checkbox? If so, what happens in #1, and what different in #2? Commented Nov 11, 2022 at 15:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.