Skip to content

Evidence Threshold Lab

Run the Evidence Threshold Lab MicroSim Fullscreen
Edit in the p5.js Editor

About This MicroSim

The xAPI runtime keeps noise out of the record with three non-evidence thresholds, constants in lrs-sim.js that authors reference and never change:

Constant Value Applies to Reason
HOVER_MS 600 ms discrete inspection by hover a pointer crossing a diagram enters many nodes in a few hundred milliseconds
MISCLICK_MS 250 ms a run a shorter run is a mis-click and would add zero-length rows
GLANCE_MS 1000 ms page dwell under one second on a page is a glance

This lab replays one fixed, scripted 60-second session: eight quick node hovers of 100 to 500 ms, three longer hovers of 700 to 1200 ms, a 120 ms run, a 40-second run, a 0.7-second glance and a 2.4-second look at a static chart, and one answered question. Each slider moves one threshold. Acts at or above a threshold turn green and become statements; acts below it turn gray and are filtered. The list shows the verb and object of every emitted statement, and the cost table counts two kinds of error: likely noise kept (crossings, a mis-click, a glance) and likely attention lost (long hovers, a real look). The answered statement is never affected.

The chapter is candid that 600, 250 and 1000 ms are design judgments, not values fitted to learner data. The lab lets you see what each judgment trades away.

Learning objective: The learner will justify a choice of hover, misclick and glance thresholds by observing how each changes the number and kind of statements produced from a fixed, scripted session.

Bloom's taxonomy level: Evaluate (verb: justify)

How to Use

  1. Start at the runtime defaults and count the statements (6 of 16 acts).
  2. Drag Hover threshold to 0 and watch every crossing become an interacted statement. Then drag it to 1500 and watch the long hovers disappear.
  3. Move Misclick threshold below 120 ms and Glance threshold below 700 ms or above 2400 ms, and read the cost table and notes.
  4. Hover any bar to see its duration and the rule that admitted or removed it.
  5. Check Show statements that were filtered out to see every act in the list, struck through when filtered. Reset to runtime defaults restores 600, 250 and 1000 ms.

Iframe Embed Code

You can add this MicroSim to any web page by adding this to your HTML:

1
2
3
4
<iframe src="https://dmccreary.github.io/microsims/sims/evidence-threshold-lab/main.html"
        height="772px"
        width="100%"
        scrolling="no"></iframe>

Lesson Plan

Audience

Teachers, instructional designers, learning-technology developers and learning-analytics practitioners (college undergraduate and professional development).

Duration

15-20 minutes

Prerequisites

  • The three MicroSim verbs and the evidence classes (Chapter 16)
  • The hover, misclick and glance thresholds and the reasons given for each

Activities

  1. Observe the extremes (5 min): For each slider, record the statement count at its minimum, at the default and at its maximum.
  2. Find the safe band (5 min): For this session, find the range of hover thresholds that keeps no crossing and loses no long hover. Do the same for the glance threshold.
  3. Justify (7 min): Write a short paragraph defending a set of three thresholds for a MicroSim used on touch screens by younger learners. Name one kind of evidence your choice would lose and explain why you accept that cost.

Assessment

  • The learner states the band of hover thresholds (above 500 ms, up to 700 ms) that separates this session's crossings from its long hovers.
  • The learner explains why no threshold changes the answered statement.
  • The learner's justification names both a kind of noise kept and a kind of attention lost, and acknowledges that the defaults are untested design judgments.

References

  1. Experience API - Wikipedia. The xAPI standard whose statements the thresholds filter.
  2. Signal-to-noise ratio - Wikipedia. The general idea behind keeping sub-threshold acts out of the record.
  3. Precision and recall - Wikipedia. A vocabulary for the trade-off between noise kept and attention lost.
  4. p5.js createSlider() reference - p5.js. The control used for each threshold.