Study Distribution (Quotas) with Filters

With Prolific, you can create a study in which you can balance your sample based on a given set of eligibility criteria.

For example, you may wish for your responses to be 50% male and 50% female participants, or 30% left handed and 70% right handed participants.

To do this, you would add a weightings argument to the filter. Weightings are expressed as ratios. Taking the ‘handedness’ example from above, if you wanted to distribute your study to 50% left handed participants and 50% right handed participants, you would use the following weighting:

1
{
2
    "filter_id": "handedness",
3
    "selected_values": ["0", "1"],
4
    "weightings": {
5
      "0": 1,
6
      "1": 1,
7
    }
8
}

Splitting your study into equal thirds for each response, your payload would look like this:

1
{
2
     "filter_id": "handedness",
3
     "selected_values": ["0", "1", "2"],
4
     "weightings": {
5
        "0": 1,
6
        "1": 1,
7
        "2": 1,
8
      }
9
}

Percentage integer values also work here, so if you wanted to distribute your study to 25% left handed and 75% right handed participants, you could use the following weighting:

1
{
2
  "filter_id": "handedness",
3
  "selected_values": ["0", "1"],
4
  "weightings": {
5
      "0": 25,
6
      "1": 75,
7
  }
8
}

You can specify the exact number of participants you want to recruit for each response. For example, if you wanted to recruit 10 left handed participants, 20 right handed participants and 5 people are ambidextrous, you could use the following:

1
{
2
  "filter_id": "handedness",
3
  "selected_values": ["0", "1", "2"],
4
  "weightings": {
5
      "0": 10,
6
      "1": 20,
7
      "2": 5
8
  }
9
}

You can specify more than one weighting requirement on a single study, and your participants will be grouped by the combination of each weighting filter. For example, the request below would create a study where you would receive 25% left handed women, 25% right handed women, 25% left handed men and 25% right handed men.

1
{
2
     "filter_id": "handedness",
3
     "selected_values": ["0", "1"],
4
     "weightings": {
5
          "0": 1,
6
          "1": 1,
7
     },
8
},
9
{
10
     "filter_id": "sex",
11
     "selected_values": ["0", "1"],
12
     "weightings": {
13
          "0": 1,
14
          "1": 1,
15
     },
16
},

Please note, study distribution with weightings is only supported for select type filters and age range filters, weightings cannot be added to other types of range filters.

Age range filters are the only range filters that support weightings, studies created with age range weighted filters can only be created, modified and published through our API. They are not yet supported in our web application.

You can specify an age range filter as follows:

1
{
2
  "filter_id": "age",
3
  "selected_range": {
4
    "lower": 18,
5
    "upper": 50,
6
  },
7
  "weightings": {
8
    "1": {
9
      "selected_range": {
10
        "lower": 18,
11
        "upper": 25,
12
      },
13
      "weighting": 1,
14
    },
15
    "2": {
16
      "selected_range": {
17
        "lower": 26,
18
        "upper": 35,
19
      },
20
      "weighting": 1,
21
    },
22
    "3": {
23
      "selected_range": {
24
        "lower": 36,
25
        "upper": 50,
26
      },
27
      "weighting": 1,
28
    }
29
  }
30
}

To see an example of using weightings filters in a study, please see the request samples area of Studies.