{{ extends "global/Page.html" }}
{{ block content }}
<h3>Instructions: Part 1 and Part 2</h3>
<ul>
  <li>You will be presented with data produced by different <em>machines</em>.
      <br>For each machine, we will show you <b>27 trials</b> generated by that machine.</li>
<span style="display:block; height:6px;"></span>

  <li>Although trials are produced by the same machine, all <b>trials are independent</b>; i.e. previous trials do not affect the next trial.
<span style="display:block; height:6px;"></span>

  <li>In every trial:
    <ul>
      <li>the machine either makes a sound or doesn’t make a sound.</li>
    </ul>
  </li>
<span style="display:block; height:6px;"></span>

  <li>A machine has lights of multiple colors, including always one <span style="color:red;">red</span>, one <span style="color:blue;">blue</span>, and one <span style="color:green;">green</span> light. In every trial:
    <ul>
      <li>you can see if the <span style="color:red;">red</span> light is on or off;</li>
      <li>you can see if the <span style="color:blue;">blue</span> light is on or off;</li>
      <li>you can see if the <span style="color:green;">green</span> light is on or off;</li>
      <li>you cannot see the other lights, which could be on or off.</li>
    </ul>
  </li>
<span style="display:block; height:6px;"></span>

  <li>The lights and the sounds may or may not be related to each other.</li>
<span style="display:block; height:6px;"></span>

  <li>That is, a trial includes information for lights that can be observed (<span style="color:red;">red</span>, <span style="color:blue;">blue</span>, and <span style="color:green;">green</span>) and on whether the machine made a sound or not. It does not provide information for lights that cannot be observed.</li>
<span style="display:block; height:6px;"></span>

  <li>Your task in Part 1 is to take summary notes (at most 75 characters) for each machine because in Part 2 you will not have access to the data observed in Part 1.</li>
<span style="display:block; height:6px;"></span>

  <li>In Part 2 you will face the same machine again and will make predictions.</li>
<span style="display:block; height:6px;"></span>

  <li>In the prediction task, you may be shown partial information about a machine's trial. This means that you may observe nothing, only the <span style="color:blue;">blue</span> light, only the <span style="color:red;">red</span> light, only the <span style="color:green;">green</span> light, or only the sound; or that you may observe two or more out of the four (that is, you may observe some lights or some lights and the sound). Your task will be to predict the elements in the trial that you are not provided with. That is:
    <ul>
      <li>If you do not receive information on the <span style="color:red;">red</span> light, you may have to predict it.</li>
      <li>If you do not receive information on the <span style="color:blue;">blue</span> light, you may have to predict it.</li>
      <li>If you do not receive information on the <span style="color:green;">green</span> light, you may have to predict it.</li>
      <li>If you do not receive information on the sound, you may have to predict it.</li>
    </ul>
  </li>
<span style="display:block; height:6px;"></span>

  <li>For payment, we will randomly select one of your predictions. If your prediction is correct, we will add {{ C.CURRENCY }}25 to your payoffs.</li>
  <li>For payment, we will randomly select one of the machines and pay 0.50{{ C.CURRENCY }} for every correct prediction.</li>
</ul>


{{ next_button }}

<br>
<div style="display:block; font-size: small;">decide: 1. keep ambiguity of what to predict? (Instructions1.2) 2. incentives. 3. make clear stochasticity? (Instructions1.2) </div>
{{endblock}}