Get started
Helping guides
Reviewing the literature can be complex—sometimes you need a helping hand, want to follow best practices to minimize bias, or simply wish to share your work. In Silvi, you don’t have to go it alone. Just invite collaborators to your project and work together seamlessly.
Inviting Collaborators ✉️
To share a project, invite other users through the project settings.
Open the project settings.
Navigate to the Collaborators tab to see an overview of existing members. Click "Add users" to invite new collaborators by entering their email.
Here you can give users the roles such as Reviewer and Resolver.
Once a user has been invited, they will see the project in their Silvi project overview.
Screening together ✅
In collaborative screening, multiple users can review studies and make independent decisions.
For example:
The user Daniel Eriksen includes a study
Another user Marie Laurent also decides to include it.
Since both agree, the study is included.
However, if they disagree, a conflict arises and would have to be resolved.
In the default setting, a study is screened as soon as a single reviewer makes a decision. However, when doing Systematic Literature Reviews (SLRs) or wants to do unbiased screening, you can require multiple reviewers to review a study before it is sceened.
Resolving conflicts ❗️
Conflicts occur when reviewers disagree. If Daniel includes a study and Marie excludes it, a conflict is flagged. Conflicts are indicated in the study overview, and the decision log with an exclamation mark.
While it is obvious that an inclusion and an exclusion is a conflict, but what if two reviewers exclude a study but for different reasons?
Example: Daniel excludes a study due to "Wrong population" while Marie excludes it for "Wrong intervention"
Whether this counts as a conflict depends on your project settings. You can configure Silvi to treat exclusion criteria as conflicts if necessary.
When there is a conflict, then only users that are resolvers can make the final decision of how the study should be screened. This is indicated to the resolver when they are about to make a decision.
Requiring multiple reviewers 👥
If you are doing a Systematic Literature Review (SLR) or simply want to ensure that your review is as unbiased as possible you can require a number of reviewers required when screening. This will be set in the project settings.
If a project requires two reviewers, a study remains unscreened until at least two reviewers make independent decisions. This is reflected in the decision log.
If the two reviewers disagree, the conflict must be resolved by a resolver.
Assigning screening tasks 📋
You can assign screening responsibilities based on project roles.
Example scenario
Three users collaborate on a project and want to distribute the screening load between each other.
Daniel Eriksen is the lead PhD student and should screen all the studies.
Professor Anna Müller is supervising the project and should review at least 30% of the studies to ensure consistency.
Marie Laurent is a research assistant and will screen the remaining studies.
In this case
Assign 100% of studies to Daniel
Assign 30% to Prof. Müller
Leave Marie unassigned - she will only screen the unscreened studies
This means that Daniel is assigned to all studies and Prof. Müller is assigned randomly to 30%. When a study is assigned to both Daniel and Prof. Müller it will remain unscreened until exactly Daniel and Prof. Müller has made a decision on the study.
Assigning studies in this scenario means that studies assigned to Daniel and Prof. Müller won’t be screened until both of them review it. If Daniel and Prof. Müller agree, the study is included and if there is a disagreement, the conflict must be resolved.
Blinding 👁️
To prevent bias, reviewers should not see each other's decisions. In Silvi, you can configure blinding settings:
Blinded projects: The decisions are not visible
Unblinded projects: Decisions are visible to all.
Blinding is set in the project settings. Here you can also configure selective blinding so that e.g., reviewers are blinded, but resolvers can see all decisions.
Feedback