Create Longitudinal Studies

EgoWeb allows you to create and administer longitudinal studies—referred to as multi-session studies in EgoWeb. This section describes how to create a multi-session study in EgoWeb.

The first step is to create a separate single-session study for each session that you want to include in the multi-session study (baseline, first follow-up, second follow-up, etc.). In each single-session study, you will program in the questions you want to ask.

(NOTE: In earlier versions of EgoWeb, each question in a multi-session study needed unique names. This is no longer true. Questions in multi-session studies can now have the same name. EgoWeb now references variables in expressions, etc. by the study name + the variable name, so re-using the variable name across linked multi-session studies is no longer a problem. Just be sure to name your study files differently to make sure you can identify them.)

The second step in creating a multi-session study is to link the single-session studies together. In the “Ego ID Question” section of each study, create a question asking for the respondent’s Ego ID, and give this question the same name across the different studies. This name should be unique to your multi-session study to ensure that EgoWeb will link together the right set of studies. In the screenshot below, this Ego ID question has the name “SRHT_ID_BF” at the baseline study. The Ego ID question in all subsequent studies that are part of this multi-session study should therefore also be named “SRHT_ID_BF.”

Advanced multi-session create egoid var.png
Creating a linking variable in the EgoID screen

After creating your Ego ID question for each of the studies you want to link together, go into the study settings for each study and select this question in the multi-session drop-down menu. This will link the studies together into one multi-session study. See the screen shot below.

NOTE: Even if you only have one study, it will show up as a multi-session study as soon as you set the Multi-session variable.

Advanced multi-session Link egoid var.png
Choosing the linking variable for multi-session studies

Once you’ve linked your studies together, you can create expressions in one study based on answers from another study in the multi-session study. So, for example, if you have a question called "gender" in your baseline instrument (called "Baseline_Study"), if you want to create an expression at the next interview study (called "Followup_Study) so that the gender question will be skipped if it was already answered by that respondent during the "Baseline_Study" interview, you can go to the expressions section of the first follow-up study and create an expression by selecting the question "Baseline_Study: gender" from the dropdown menu.

When interviewing, if the EgoID variable entered has been used in another study session, the data will be linked in the new interview.

Web-based Studies with Invited Participants

NOTE: This feature works for server installations of EgoWeb. If you have an EgoWeb installation url that starts with "localhost" you have a PC or Mac install and will not be able to send these links to participants.

There are currently two ways of starting an interview. First, interviews can be started by logging in to EgoWeb and using the “Interviewing” link to start a new interview. This requires a user account to be added to “User Admin”. Users can be categorized as Interviewers and assigned to specific studies. This limits these accounts to only conducting interviews on studies assigned to them. However, this does not work for studies in which you want participants to fill out their own responses and not be added to the EgoWeb server as a user.

The second way to initiate an EgoWeb interview is through creating a study with Pre-Defined Participants. This allows you to invite users to fill out their own responses without having them be assigned EgoWeb users.

Step 1: Add Pre-defined Participants
pre-def participant.jpg
Pre-Defined Participant Entry

.csv file with participant names and email addresses must be in this format to be imported correctly. DO NOT import an excel file. Save an excel file as a .csv format before importing. Do not include column headings.

pre-def participant import csv.jpg
Format of .csv file for Pre-defined Participant Importing.

Step 2: Create EgoID question for capturing authenticated email associated with link.

EgoID screen.jpg
EgoId Screen for linking email to url for Pre-Defined Participant

Step 3. Generate links for participants. There are two ways to generate links. The first and most time intensive way is to manually click on the link icon to create a link to send to each participant. The second way is to export the entire Pre-Defined Participant data into a .csv file. This file can be used in email programs to send out invitations to a whole list of participants at one time with the associated url. To generate this file, first, go to Data Processing for the study with Pre-Defined Participant Data. Click on the "Export Pre-defined Alter List" button. (NOTE: the button incorrectly says "Alter" but it is for the Participants. The name has not been changed from a previous version of this procedure.)

pre-def participant export.jpg
Export Pre-defined Participant List

pre-def participant export csv.jpg
.csv file with exported pre-defined participant data associated links

Dyad Match

Dyad Match provides an interface to do the alter matching described in this paper: The Analysis of Duocentric Social Networks: A Primer
The interface provides a way to identify if two respondents in two separate EgoWeb interviews named the same alters. Matched alters are flagged in the ego-alter data export for analysis in other software.

This interface is still in progress. It is going to be used for a new study of low income newlywed couples. Most of the programming has been completed but it is still being tested. Check back later for more detailed instructions.

Return to the EgoWeb 2.0 Wikispaces Home