Discussion 4

Viewing 7 reply threads
  • Author
    Posts
    • #7037
      Liz Lewis
      Moderator
      • Have you encountered many/any box-and-whisker plots in the course of your work? On other dashboards?

      • What about histograms?

      • If you have little or no prior experience with them, do you think they are easy to understand? If not, can you think of some alternate ways to display distribution data?

    • #7112
      Kara Showers
      Participant

      Have you encountered many/any box-and-whisker plots in the course of your work? On other dashboards?
      I have only encountered them rarely.

      What about histograms? A few.

      If you have little or no prior experience with them, do you think they are easy to understand? Not really at first glance.

      If not, can you think of some alternate ways to display distribution data? The bar graphs that have bars above & below a certain point (i.e. a median). Maybe scatter plots?

    • #7113
      Ian Lemieux
      Participant

      >> Have you encountered many/any box-and-whisker plots in the course of your work? On other dashboards?

      I’ve encountered box-and-whisker plots in my work, but to be entirely honest, have always had a difficult time deciphering what’s what. I’ve found the leading diagram on this site, with explanations, particularly helpful and a useful reference/reminder:
      https://flowingdata.com/wp-content/uploads/2008/02/box-plot-explained.gif
      As for where, most commonly I’ve seen box-and-whisker plots used in presentations of clinical trial data in study reports, manuscripts, and medical journals. Closest I’ve seen to use within a dashboard, although not a dashboard, has been in abstracts and/or on research posters.

      >> What about histograms?

      Until this course, I’d not tracked the difference between a histogram and a bar graph. Imagine I’ve seen them, yet not known they were their own entity.

      >> If you have little or no prior experience with them, do you think they are easy to understand? If not, can you think of some alternate ways to display distribution data?

      Box-and-whisker plots are likely tough to follow, especially for the first-timers or readers who are familiar, but have less experience with them. That said, I think histograms do a nice job of presenting distribution data, albeit with fewer criteria. Side-by-side histograms, as a rough means of comparison could be leveraged. If too cumbersome and/or there is not enough space, that’s where box-and-whisker plots allow you the flexibility of presenting distribution data in parallel for ease of comparison.

    • #7123
      Liz Lewis
      Moderator

      Great comments so far, Kara and Ian. I’ll see if I can find a data set that we used in grad school. If I recall correctly, the average (mean) was the same for the three sub-populations, as were some of the other metrics. But the shape of the data in each cohort was quite distinct, and my recollection is that one of these types of graphs really helped you to see why and where the differences occurred. I’ll let you know what I find.

    • #7125
      Masis Parunyan
      Participant

      I’ve mainly encountered box-and-whisker plots being used to display results from scientific experiments in reports and papers. I have not seen them used on other dashboards. The histograms I have seen have mainly been in the setting of displaying grade distributions on examinations.

      I think histograms are easy to understand at a glance. Box-and-whisker plots, however, are less immediately clear and require an explanation of what the box and “whiskers” represent. I think that once you know how to read them, they are useful for showing more information about the statistical distribution of the data.

    • #7133
      Allyson Cogan
      Participant

      *Have you encountered many/any box-and-whisker plots in the course of your work? On other dashboards?
      I have rarely seen box and whisker plots used and the few times I did encounter them in a few peer reviewed journals I found them hard to decipher and moved on to data representations that were more intuitive.

      *What about histograms?
      I would say I have seen histograms used more frequently than box and whisker plots but also must admit I believe in hindsight I simply considered them bar charts with different formatting.

      *If you have little or no prior experience with them, do you think they are easy to understand? If not, can you think of some alternate ways to display distribution data?
      I find histograms more user friendly and they seem to portray the important data more clearly as well as prominently. I can certainly see the use of box and whisker charts but they definitely require prior experience concerning best practices for deciphering the data they are presenting. Either way they are both effective means of illustrating distribution data as long as the individuals reading the data is familiar with them.

    • #7136
      Sherry Roberts
      Participant

      Have you encountered many/any box-and-whisker plots in the course of your work? On other dashboards?
      Answer: We see box-and-whisker a lot at work, clinical trials use these types of displays often to show statistical outputs of data. It help to visualize the norms and outliers in datasets. I haven’t really seen this type of visualization in other dashboards that I have encountered.

      What about histograms?
      Answer: To be honest I had not appreciated the difference between a histogram vs a bar chart before this course. It’s great to learn something new. I am wondering if I have been looking at histograms but not understanding what they were in the past.

      If you have little or no prior experience with them, do you think they are easy to understand? If not, can you think of some alternate ways to display distribution data?
      Answer: I have always struggled with box-and-whisker before, the class exercises were helpful because in the work setting where I normally see them it is assumed that we all understand them. Ian’s clip that he provided was very helpful also. I have a much better understanding of what I am looking at now! Thank you!

      I’m not sure about the histogram, I will be more aware of the difference going forward so I think time will tell how clear histograms are to me. Stay tuned!

    • #7138
      Emilia Cain
      Participant

      Have you encountered many/any box-and-whisker plots in the course of your work? On other dashboards?
      I´ve been working in public health for almost 7 years and I have rarely seen a box-and-whisker plot. I´ve seen them in journals buy not really for work purposes. I think their understanding is not very straight forward and maybe that´s why not commonly used.

      What about histograms?
      Yes, I do see a lot of histograms.

      If you have little or no prior experience with them, do you think they are easy to understand? If not, can you think of some alternate ways to display distribution data?
      I think they are easy to understand if you have some basic knowledge on their interpretation; otherwise, they are not easily understood. Where I work, most of the information generated is intended to be used by the public or other stakeholders who are not necessarily trained in public health and therefore, we don´t use them often.

Viewing 7 reply threads
  • You must be logged in to reply to this topic.