Item criteria based on response in another project

Hi all,

My questions is about the opportunities for item criteria. In the Ethica learning page it is stated that that criteria can be applied to answers of previous questions within the same survey (i.e., Q{item}), or an item from a different survey within the same study (i.e., Q{survey}_{item}). I would like to know whether there are values from other sources that are eligible for this criteria option as well. More specifically, is it possible to display an item based on the activation of a trigger (e.g. display item only for the first time trigger). Is it possible to take values from a survey in a different study/ project as a criterium?

I am looking forward to your suggestions!

Best,
Teun

Basically, I would like to create two projects: one with a baseline survey (let’s say survey_ID = x), and another with a main survey. I want them to be in separate projects because of relative time triggering concerns. In the baseline survey, we want participants to indicate in which class they are (e.g. Q1, with 4 options), and depending on the response, we would like to present them specific questions.

So, when I want to present the question in the main survey only for the class that we said to be answer 1, I assume the criteria for the question in the main survey should be:

Qx_1 == 1

Am I correct? Because, unfortunately, this does not work. Therefore, I wonder whether it is possible to use answer responses of surveys in project 1 as a criteria for items of surveys in project 2.

Hopefully, someone can help me out.

-Teun

Hi @t.siebers

First, sorry for the delay in answering your question.

Unfortunately, it’s not possible to use responses in one project for skip patterns and branching in another project. In Ethica, projects (or Studies) are completely isolated and you cannot reference each other.

That being said, each study in Ethica offers quite a few options and features, and I cannot think of any scenario where having two separate Ethica studies is necessary for the same real-world study. Can you explain why you need to create two separate studies? I might be able to offer an alternative solution.

Thank you,
Mohammad

Hi @m.hashemian,

Thanks for your reply. The reason why we want to set up two studies is because we have a baseline measurement, which starts a few weeks before the main study. We do not impose restrictions on the time of registration for the baseline session, meaning that they can sign up any time before a specific date. However, the triggers in the main study should be at a fixed date and time. Since triggers can only be set up relative to the time of registration, we do have a problem if we put the baseline and the main survey in the same project.

Our backup plan is to avoid branching by creating 20+ different projects, but that would not be a very elegant solution…

I would be happy to hear about any other ways to solve this problem!

Best,
Teun

One way to do so is to use absolute dates for surveys instead of relative to registration time. Let’s say you want the study to start on Dec. 1st, and finish on Dec. 21st, but you want want to allow people to join anytime before Dec. 1st.

You set the start time of the study to now, and end time to Dec. 21st. For surveys, you specify to be prompted on Dec. 1st, 2019 with the repeat schedule that you intend (this is not possible via the Survey Editor yet, but the feature is available in Ethica). The only survey that is relative to the registration time is the baseline. So when a participant joins say on Nov. 10th, they get the baseline survey right away, but the other surveys are started from Dec. 1st. If someone joins on Nov. 15th also, she will get the baseline right then, and the other surveys from Dec. 1st as well.

One catch here is that if your study includes collecting sensor data, using this method means that sensor-based data collection starts as soon as the participant starts the study (e.g. Nov. 10th and Nov. 15th in the above examples), rather than on Dec. 1st.

That sounds perfect! Unfortunately, I have not been able to find the absolute time triggers (e.g. setting it to 10:00:00 on December 1st, 2019) in Ethica… Could you tell me where to find this option? Can I do this with JSON code?

-Teun

Unfortunately (or, fortunately, if you are comfortable with JSON!), this can only be done in JSON. To do so, create your surveys and all triggering logics, then download the JSON file and open it with a text editor. You would need to format the JSON text in order to be able to read it easier. Then you see the content of your survey like the following:

{
    "version": 1,
    "name": "....",
    "triggering_logics": [
        {
            "notifications": [
                0
            ],
            "type": "time",
            "id": 0,
            "schedule": {
                "time_format": "relative",
                "base": "registration_date",
                "end": {
                    "criteria": "never"
                },
                "period": {
                    "start": "0 14:30:00",
                    "randomization": {
                        "function": "uniform"
                    },
                    "end": "0 14:30:00"
                },
                "repeat": {
                    "week_days": [
                        "mon",
                        "tue",
                        "wed",
                        "thu",
                        "fri",
                        "sat",
                        "sun"
                    ],
                    "occurrence_unit": "week",
                    "occur_every": 1
                }
            }
        },
    ],
    "notifications": [
        ...
    ],
    "sub_surveys": [
        ...
    ],
    ...
}

Here you see the start and end time of your survey is defined in relative time ("start": "0 14:30:00"). Modify that to an absolute time, like 2019-12-01 00:00:00:

{
    "version": 1,
    "name": "....",
    "triggering_logics": [
        {
            "notifications": [
                0
            ],
            "type": "time",
            "id": 0,
            "schedule": {
                "time_format": "relative",
                "base": "registration_date",
                "end": {
                    "criteria": "never"
                },
                "period": {
                    "start": "2019-12-01 14:30:00",
                    "randomization": {
                        "function": "uniform"
                    },
                    "end": "2019-12-01 14:30:00"
                },
                "repeat": {
                    "week_days": [
                        "mon",
                        "tue",
                        "wed",
                        "thu",
                        "fri",
                        "sat",
                        "sun"
                    ],
                    "occurrence_unit": "week",
                    "occur_every": 1
                }
            }
        },
    ],
    "notifications": [
        ...
    ],
    "sub_surveys": [
        ...
    ],
    ...
}

Then save the json file and upload it again.

Note that in this case, the Survey Editor does not properly show the new timing of your triggering logic, but when a participant joins you should be able to see the proper time in the Survey Sessions page of the Researcher Dashboard.

Hope this helps

Also, if you are comfortable reading and modifying JSON files, you can find the detailed description of Ethica Survey’s JSON template in the survey schema.

That’s exactly what I was looking for, thank you so much!

Teun

1 Like

A post was split to a new topic: Configuring study schedule