Job: extend your experiment

• instead of pictures (of cats etc.), use automatically generated pictures of colored shapes, like these:

  

• do not generate pictures manually beforehand, but generate a random picture on each trial

Job description

use the CanvasTemplate, available here: https://github.com/babe-project/CanvasTemplate

• adapt it to do the following, on each trial:
  • sample total_set_size from {10,25,50} and a non-zero focal_set_size of at most total_set_size
  • choose a random shape from {circle, square, triangle}
  • display focal_set_size of shape in the focal_color, the rest in the other_color (choose colors from {blue, red, yellow, green})
  • arrange colored shapes arbitrarily on the screen
  • dynamically construct the sentence to be judged/rated as: “Some of the shapes are focal_color.”
  • record all relevant random choices for each trial’s picture

Canvas template

• in index.html:

<div class="stimulus-container">
    <canvas id="canvas" height="400" width="600"></canvas>
</div>

• in helpers.js:

var drawOnCanvas = function(canvasElem, trialInfo) {
    var canvas = createCanvas(document.getElementById('canvas')); // defined in scripts/canvas.js
    var coords = canvas.getCoords(trialInfo.total, trialInfo.size);
    for (var i=0; i<trialInfo.total; i++) {
        if (i < trialInfo.focalNumber) { canvas.draw(trialInfo.focalShape, trialInfo.size, coords[i].x, coords[i].y, trialInfo.focalColor); } 
        else { canvas.draw(trialInfo.otherShape, trialInfo.size, coords[i].x, coords[i].y, trialInfo.otherColor); }
    }
};

• in views.js:

drawOnCanvas(document.getElementById('canvas'), exp.trial_info.main_trials[CT]);

Rmarkdown: overview

• lean markup language which allows execution of R code
  • documentation: https://rmarkdown.rstudio.com
• exports to multiple formats
  • simplifies a reproducible work flow

APA guidelines to structure experiment reports (roughly)

• (optional) motivation/design
  • what kind of experiment (e.g., between vs. within-subject design) and why
• participants
  • how many participants were recruited how/from where; gender distribution; age; payment etc.
• materials & apparatus
  • which pictures/sentences/instructions/machinery was used (e.g., which eye-tracker, …)
• procedure
  • how the experiment was structured, how many trials of which kind in which order etc.
• (optional) analysis (also sometimes referred to as ‘design’)
  • what are the dependent and independent variables used for statistical analysis
• (optional) data preparation
  • data manipulation before further analyses, e.g., data cleaning, variable transformation
• results
  • summary statistics (e.g., mean error rates), plots & analyses (e.g., \(p\)-values)
• discussion
  • interpret the results