Differences

This shows you the differences between two versions of the page.

--- wiki:autolit:admin:configure [2024/05/22 11:47]
jthurnham [2. Synthesis]
+++ wiki:autolit:admin:configure [2024/09/23 11:54] (current)
jthurnham [Nest Access]
@@ Line 16: / Line 16: @@
   * **Exception: **if you are giving an Organization access to or ownership of a nest, that is a Nest-level Admin function; see the [[:wiki:autolit:access|Manage Access page]].
-{{ :undefined:screenshot_2023-12-06_at_15.39.12.png?nolink |}}
+{{ :undefined:screenshot_2024-09-23_at_12.42.09.png?nolink |}}
-==== Tag Recommendations ====
+===== 2. Synthesis =====
-Standard tag recommendations are available to all users to assist in the Tagging process. Smart Tag recommendations are only available to enterprise users, but utilize OpenAI GPT 3.5/4 to search Abstracts and Full Texts. Toggling on "Smart (Beta)" will automatically generate recommendations for any included studies with available full texts. This will run in the background.
+==== Outputs Displayed ====
-You have a limited number of regenerations due to the costliness of each generation. After full text smart recommendations are generated, and for further assistance, you have the option to also generate smart tag recommendations for included abstracts. Learn more about [[:wiki:autolit:tagging:tag_recommendations|Tag Recommendations. ]]
+From the Synthesis section, you can toggle ‘on’ or ‘off’ any Synthesis page: Qualitative Synthesis (QLS), Quantitative Synthesis (QNS), Manuscript, Dashboard, Critical Appraisal, or PRISMA.
-==== Import Existing Tags ====
-If there are tags from a previous nest that are relevant to your current nest, you can [[:wiki:autolit:admin:import#import_tags|import existing tags]] for any record already in your nest by uploading a spreadsheet. To do this, under Tagging on the Settings page, selecting “Begin” under Import.
+{{ :undefined:screenshot-2024-09-05-at-12.32.00-2048x352.png?nolink |}}
-----
+By default:
-===== 5. Meta-Analytical Extraction (previously named Extraction) =====
+  * QLS and PRISMA outputs are toggled on
+  * QNS and Critical Appraisal are disabled (this is because QNS is generated from the Meta-Analytical Extraction module, and Critical Appraisal outputs are generated from the Critical Appraisal module, which are both off by default; if you’d like to toggle these on in Synthesis, first turn on and complete MA Extraction and/or Critical Appraisal further down the Settings page, then return to toggle these on in Synthesis)
+  * Manuscript and Dashboard are toggled off (currently, only Manuscript or Dashboard can be toggled on at the same time but all work is saved when toggled between; the corresponding Manuscript Editor or Dashboard Editor will then be displayed under the left-hand Synthesis menu item)
-==== MA Extraction Modes: Standard vs. Dual ====
+{{ :undefined:screenshot-2024-09-05-at-12.43.07-2048x370.png?nolink |}}
-As an Admin, you can configure the Meta-Analytical Extraction Mode to be:
+==== Default QLS Display Type ====
-  * Standard (default)
+Choose which QLS diagram is displayed when you open the Synthesis page: Sunburst or Dendrogram.
-  * [[:wiki:autolit:meta_analytical_extraction:dual|]]
-**Standard MA Extraction**  means that each record will be extracted for quantitative data only once.
+{{ :undefined:screenshot-2024-09-05-at-12.45.37.png?nolink |}}
-**Dual MA Extraction**  means that each record will be extracted by two different users and adjudicated by a third. To toggle Dual MA Extraction 'on', select it in the Admin page:
+By default, the Sunburst diagram is selected and displayed in Synthesis:
-{{  :undefined:settingsdualma.png?nolink&  }}
+{{ :undefined:screenshot-2024-09-05-at-12.47.43-2048x1041.png?nolink |}}
-{{https://wiki.nested-knowledge.com//home/ubuntu/nkwiki/dokuwiki/lib/plugins/ckgedit/fckeditor/userfiles/image/wiki/autolit/fourteen.png?nolink&}}
+But if the Dendrogram is selected, it will be displayed in Synthesis:
-===== 6. Critical Appraisal (previously named Risk of Bias) =====
+{{ :undefined:screenshot-2024-09-05-at-12.48.49-2048x1046.png?nolink |}}
-Choose a Critical Appraisal system. Learn more about [[:wiki:autolit:critical_appraisal#configure_critical_appraisal|Configuring Critical Appraisal]]
+Note: both diagrams may still be accessed at anytime (circled in red) but this Setting decides which is shown by default.
-{{  :wiki:autolit:admin:img7.png?nolink&  }}
+==== Nest Access ====
-----
+  * **Make a nest public.**  Toggle this on to make the link to your [[:wiki:synthesis|]] page (and all sub-pages) publicly viewable.
+  * **Make a nest protected.**  Toggle this on to make the link to Synthesis undiscoverable and inaccessible to those without the key (or explicitly granted access). You will still be able to share a link with external parties, and the external party will not have to log in to view Synthesis. In this mode, you may also choose to hide the ability to share the outputs from the Synthesis for any users viewing it including yourself.
+  * **Make a nest private.**  Toggle this on for the most protection. Only those explicitly granted access with a NK account will be able to view the Synthesis.
-===== 7. Blinding =====
+{{ :undefined:screenshot-2024-09-05-at-12.51.06.png?nolink |}}
-To avoid bias in your review where multiple reviewers are participating, you may want to consider blinding certain aspects of the process. The Blinding settings allow you turn off or hide features.
+==== Show Full Texts ====
-**Blind Adjudication**  allows you hide the reviewer's identity for each decision as to not influence the ultimate decision. For example:
+Abstracts are automatically shown alongside records displayed in Synthesis. If you wish to also display Full Texts, toggle this on here. Due to copyright reasons, this cannot be turned on if your nest is public, only protected or private (see above).
-Adjudicate Screening before Blinding:
+{{ :undefined:screenshot_2024-09-23_at_12.46.34.png?nolink |}}
-{{  :undefined:blind.png?nolink&  }}
+==== Static Manuscript ====
-After Blinding:
+Upload a pdf to be shown in Synthesis when Manuscript is toggled on and empty.
-{{  :undefined:blind2.png?nolink&  }}
+{{ :undefined:screenshot-2024-09-05-at-12.54.05.png?nolink |}}
-**Hiding Inclusion Model**  allows you to remove the inclusion probability for each study if turned on, so the reviewer can make a completely independent decision.
+Manuscript in Synthesis will then display the pdf as such:
-Screening before hiding probabilities:
+{{ :undefined:screenshot-2024-09-05-at-12.55.35-2048x1052.png?nolink |}}
-{{  :undefined:hide.png?nolink&  }}
-After hiding probabilities:
-{{  :undefined:hide2.png?nolink&  }}
-**Hiding Comments**  removes the Comments tab in the workflow modules (Screening, Tagging, MA Extraction, Critical Appraisal), allowing you to hide any comments made on studies when viewing them. As a result comments can't directly be made as they appear in the queue but can be made and seen when the study is accessed in Study Inspector.
-Right hand side tabs before hiding comments:
-{{  :undefined:comments.png?nolink&  }}
-After hiding comments:
-{{  :undefined:commentsno.png?nolink&  }}
-Hiding Tagging During Screening removes the Tagging tab available as you screen. We offer this capability during screening to be able to apply preliminary tags, but similarly it also exposes users to bias after a previous reviewer has screened and tagged. The before and after mirrors the above Comments removal but for the Tagging tab instead in Screening alone.
-===== 8. Copying, Renaming, or Deleting a Nest =====
-Copying a nest will copy the data and configurations, such as the search terms, references, exclusion reasons, the tagging hierarchy, applied tags, data elements, full-text PDFs, Critical Appraisal systems, the protocol and manuscript, and other configurations. The copied nest is entirely independent of the original nest. Access privileges will not be copied from the old nest and must be configured.
-**Only Owners have the right to delete a nest.**  "Delete" can be found next to "Rename"; deleting a nest is irreversible and will completely delete all data related to the project in question!
-{{  :undefined:screenshot_2023-12-06_at_15.42.05.png?nolink&  }}
-<WRAP center round alert 60%> \\ Deleting a nest has wide-reaching and irreversible effects, including loss of all nest data. Proceed with caution! \\ </WRAP>
 ----
-===== Enterprise-level Features =====
-These features are only available to admins within an organization subscribed at the Enterprise level.
-==== 1. Allocation ====
-Allocation allows you to assign work in specific modules to various members of your team. Learn more about [[:wiki:autolit:admin:allocation|]].
-==== 2. Nest Backups ====
-Nest Backups allow you to manually save up a backup of your nest to capture a version of it at a specific date and time. Learn more about [[:wiki:autolit:admin:nestbackups|]].
 ===== 3. Screening =====
@@ Line 143: / Line 104: @@
 ==== Screening Model ====
-**Manual vs. Automatic:**  By default, the Screening Model will be on Manual, meaning that it will only run when a user selects Train Screening Model during the Screening process. Toggling the Screening Model on (see image above) will alter this so that the Model runs as soon as a sufficient number of studies are screened, and this Model will be re-run automatically as more screening occurs. In the image below, if manual the red box will say "Train Screening Model," and if automatic training is on and enough studies are screened, the inclusion probability will be provided– in this case it is "P(Inclusion): 0.87."
+**Manual vs. Automatic:**  By default, the Screening Model will be on Manual, meaning that it will only run when a user selects Train Screening Model during the Screening process.
+{{ :undefined:screenshot_2024-07-24_at_17.27.20.png?nolink |}}
+For Automatic Training, head to Nest Settings--> Screening Model. This will cause the Model to run as soon as a sufficient number of studies are screened, and this Model will be re-run automatically as more screening occurs.
+{{ :undefined:screenshot_2024-07-24_at_17.25.54.png?nolink |}}
+As probabilities are generated, they are displayed in place of the "Train Screening Model" text in the Screening module, in the below case it is "P(Inclusion): 0.87.".
+{{  :undefined:2screen.png?nolink&  }}
 **Displayed vs. Hidden:**  By default, the Screening Model will be displayed, meaning that any user will be able to see the probabilities on the Screening panel:
@@ Line 149: / Line 120: @@
 {{  :undefined:2screen.png?nolink&  }}
-If you want this probability to be hidden from all users, Admins, and Owners on each individual study, toggle "Hide Probabilities." If hidden, you can still run the model from the Admin page and have the option to turn on Automatic updating. [[:wiki:autolit:screening:inclusionpredictionmodel|Learn more about the Screening model here.]]
+If you want this probability to be hidden from all users, Admins, and Owners on each individual study, toggle "Hide Probabilities." If hidden, you can still run the model from the Admin page and have the option to turn on Automatic updating. [[:wiki:autolit:screening:model|Learn more about the Screening model here.]]
 ==== Import Existing Screening Decisions ====
@@ Line 194: / Line 165: @@
 ===== 6. Critical Appraisal (previously named Risk of Bias) =====
-Choose a Critical Appraisal system. Learn more about [[:wiki:autolit:critical_appraisal#configure_critical_appraisal|Configuring Critical Appraisal]]
+Choose a Critical Appraisal system. Learn more about [[:wiki:autolit:critical_appraisal#configure_critical_appraisal|Configuring Critical Appraisal.]]
-{{  :wiki:autolit:admin:img7.png?nolink&  }}
+{{ :undefined:screenshot_2024-05-22_at_12.51.27.png?nolink |}}
 ----
@@ Line 204: / Line 175: @@
 To avoid bias in your review where multiple reviewers are participating, you may want to consider blinding certain aspects of the process. The Blinding settings allow you turn off or hide features.
-**Blind Adjudication**  allows you hide the reviewer's identity for each decision as to not influence the ultimate decision. For example:
+==== Blind Adjudication ====
+Reviewers' identities are shown to adjudicators by default. You can hide these identities for each adjudication decision as to not influence the ultimate decision. To do this, toggle on "Blind Adjudication":
+{{ :undefined:screenshot_2024-06-14_at_15.01.57.png?nolink |}}
+For example:
 Adjudicate Screening before Blinding:
@@ Line 214: / Line 191: @@
 {{  :undefined:blind2.png?nolink&  }}
-**Hiding Inclusion Model**  allows you to remove the inclusion probability for each study if turned on, so the reviewer can make a completely independent decision.
+==== Queue Ordering ====
+By default, records shown in the screening queue to reviewers is in order of Ref ID. You may wish to randomize the studies shown to avoid bias or order them by inclusion probability after the model is trained. Selecting the queue order affects the remainder of studies left in the screening queue.
+{{ :undefined:screenshot_2024-06-14_at_15.07.13.png?nolink |}}
+==== Hide Probabilities ====
+When you have trained a Screening Model either manually or via automatic training these values are displayed in Screening. To avoid bias, you may wish to hide these probabilities. You can hide these probabilities in Nest Settings. This allows you to remove the advancement or inclusion probabilities for each study if turned on, so the reviewer can make a completely independent decision.
+Note: Advancement probabilities are shown in Two-Pass Screening at the Abstract level, and Inclusion probabilities are shown in Standard Screening in the singular round of screening.
 Screening before hiding probabilities:
@@ Line 224: / Line 211: @@
 {{  :undefined:hide2.png?nolink&  }}
-**Hiding Comments**  removes the Comments tab in the workflow modules (Screening, Tagging, MA Extraction, Critical Appraisal), allowing you to hide any comments made on studies when viewing them. As a result comments can't directly be made as they appear in the queue but can be made and seen when the study is accessed in Study Inspector.
+//If Robot Screener is on, these probabilities are automatically removed.//
+==== Hide Comments ====
+This removes the right-hand Comments tab in the workflow modules (Screening, Tagging, MA Extraction, Critical Appraisal), allowing you to hide any comments made on studies when viewing them. As a result comments can't directly be made as they appear in the queue but can be made and seen when the study is accessed in Study Inspector.
 Right hand side tabs before hiding comments:
@@ Line 234: / Line 224: @@
 {{  :undefined:commentsno.png?nolink&  }}
-Hiding Tagging During Screening removes the Tagging tab available as you screen. We offer this capability during screening to be able to apply preliminary tags, but similarly it also exposes users to bias after a previous reviewer has screened and tagged. The before and after mirrors the above Comments removal but for the Tagging tab instead in Screening alone.
+==== Hide Tagging ====
+During Screening you may remove the Tagging tab available as you screen. We offer the ability to Tag during screening to be able to apply preliminary tags, but similarly it also exposes users to bias after a previous reviewer has screened and tagged. The before and after mirrors the above Comments removal but for the Tagging tab instead in Screening alone.
 ===== 8. Copying, Renaming, or Deleting a Nest =====
@@ Line 256: / Line 248: @@
 Allocation allows you to assign work in specific modules to various members of your team. Learn more about [[:wiki:autolit:admin:allocation|]].
-==== 2. Nest Backups ====
+==== 2. Nest Versions ====
-Nest Backups allow you to manually save up a backup of your nest to capture a version of it at a specific date and time. Learn more about [[:wiki:autolit:admin:nestbackups|]].
+Nest Versions allow you to manually save up a version of your nest at a specific date and time. Learn more about [[:wiki:autolit:admin:nestbackups|Nest Versions]].

Nested Knowledge

User Tools

Site Tools

Differences

Page Tools