Handling Insignificance in UX Data

Ғылым және технология

After collecting KPI numbers for two design versions, the difference between the two #metrics is not statistically significant. Now which version should you launch?
#UX

Пікірлер: 8

  • @suziross
    @suziross2 жыл бұрын

    As a UX researcher at LEGO, I love that you used LEGO bricks to show this! 😁

  • @gamma_girl
    @gamma_girl3 жыл бұрын

    Strong qualitative information should overrule weak quantitative information 👍🏽

  • @dougajmcdonald
    @dougajmcdonald3 жыл бұрын

    The takeaway I took from this is that, as always, the main risk is never being able to iterate after a release. Nothing is ever certain, and we as UX professionals should embrace that and be building a culture of continual review and improvement, as Jacob illustrates, as long as it's commercially viable to do so. Perfect is the enemy of released. But "done" is the enemy of continual improvement.

  • @karlr2908
    @karlr29082 жыл бұрын

    What a great video!

  • @Puleczech
    @Puleczech3 жыл бұрын

    Please, allow comments on other videos as well.

  • @TonyBolero
    @TonyBolero3 жыл бұрын

    Excellent explanation! Even I get it :)

  • @shantyman161
    @shantyman1613 жыл бұрын

    And don't forget: Qualitative data does not only tell you which design to use, but also why. And that might lead to completely new and even better designs.

  • @daph104
    @daph1043 жыл бұрын

    Quality.

Келесі