The Dashboard & the 4P Framework (Pull, Pick, Prep, Push) | HubMeta Tutorial #3 — Transcript

Full transcript

1. 0:14Welcome back. So, we created our very
2. 0:17first project. For this series of
3. 0:20videos, I'm going to use a project that
4. 0:22has data in it already just for you to
5. 0:26see how our dashboard looks in a live
6. 0:28project. I'm going to use this one,
7. 0:31which is a meta-analysis on the effect
8. 0:34of multinationality on performance. This
9. 0:37is an international business research.
10. 0:39And when you come here, this is the
11. 0:42dashboard that you will see. As you can
12. 0:44see, it is a still loading. It's not
13. 0:47showing any numbers. And this helps me
14. 0:50explain a concept that we have
15. 0:52implemented in the new Hub meta.
16. 0:55In this platform, to make things faster,
17. 0:58one feature that we have implemented is
18. 1:01an efficient use of cache in your
19. 1:04browser. So, a whole lot of data is
20. 1:08downloaded only once in your browser.
21. 1:11And the first time it is loading, it
22. 1:14will take, as we just watched in this
23. 1:16video, anywhere between 20 seconds to 1
24. 1:19minute depending on your internet speed
25. 1:21and how you're connecting. But once it
26. 1:24is downloaded, our browsing of the
27. 1:27articles, of going into each page,
28. 1:31becomes much, much faster because data
29. 1:34is now stored in your browser. This is
30. 1:37an important concept because in
31. 1:40systematic reviewers, we are dealing
32. 1:42with a huge body of text and papers and
33. 1:47a whole lot of that. And this is sort of
34. 1:49our compromise to make things faster for
35. 1:51you. What do I see in this dashboard
36. 1:54page is just a quick peek into how far I
37. 1:59am with the project. Like the main
38. 2:02numbers that I need, like the number of
39. 2:04articles that have been imported here,
40. 2:07how many of them have remained
41. 2:08unscreened, how many of them I have
42. 2:11included after going through the title
43. 2:14and abstract review, how many of them
44. 2:17have been included after full-text
45. 2:19review. Just these dashboards on the
46. 2:22right, like how it looks in terms of
47. 2:24numbers, progress in each of the steps,
48. 2:27as you can see, like how much progress
49. 2:30I've had in each of the steps.
50. 2:32And then, you know, a little bit of
51. 2:35breakdown
52. 2:36the number of papers in each. Down here,
53. 2:39I will see the team members that I have,
54. 2:42and I can add members from here. Here,
55. 2:45we also have a Prisma chart. This one,
56. 2:48if I click on it, it will bring the
57. 2:50Prisma chart as, you know, the
58. 2:53guidelines coming from Prisma. Citation
59. 2:56is down here. So, you can you can view
60. 2:59it here, or you can download it as a PNG
61. 3:03to put it in your paper when you're
62. 3:05reporting. Now, other things that we
63. 3:08have here is our share button. When we
64. 3:11share, you can add the email address
65. 3:14that your colleague has used to sign up
66. 3:17on HubMed. When you invite them, they
67. 3:21will be able to join this project. So,
68. 3:24you just enter their email and hit
69. 3:26invite. And what happens is that when
70. 3:30they come in and log in to their HubMed,
71. 3:34inside these invitations tab, a new
72. 3:37invitation from this new project will
73. 3:40pop up, and then, after that, they can
74. 3:42join as a collaborator to your project.
75. 3:45We also have project settings here,
76. 3:47which shows our team members. Like I
77. 3:49explained, I went out and came back
78. 3:51here, so it is loading data back into
79. 3:54our cache again. So, we are back here.
80. 3:57When you add a member, you can define
81. 3:59their role to be a reviewer or a manager
82. 4:02of the project. We will explain what
83. 4:04these are when we are going through our
84. 4:07title screening. And then, you can use
85. 4:09backups of your project. What backups
86. 4:12are is that we keep a backup of your
87. 4:14project every day. So, at the end of the
88. 4:18day, we will
89. 4:20keep a copy of this project on our
90. 4:22server. And then, this goes back about,
91. 4:26I think, 15 days, and then we will purge
92. 4:28it after that. This is just
93. 4:31because it has happened many, many times
94. 4:33in our own projects that we did
95. 4:35something that we didn't like, or one of
96. 4:38our project collaborators made a change
97. 4:40that messed things up, and we just want
98. 4:43to go back to a former version. Here, we
99. 4:47can also change the screening criteria.
100. 4:50I'll try to be brief in these videos.
101. 4:52Some of these features are things that
102. 4:53you can figure out on your own. Now, the
103. 4:55last thing I want to explain in this
104. 4:56page is the four steps of doing
105. 5:00systematic reviews. In our new platform,
106. 5:04we have created the workflow of a
107. 5:06systematic review into four main steps.
108. 5:10The first step is pull. This is where we
109. 5:13grab all the information that we want to
110. 5:16process. So, we import articles from
111. 5:19many different sources.
112. 5:21We put them into our meta. We remove
113. 5:23duplicates. We work on them, or we
114. 5:25create search keywords, transfer from an
115. 5:29older platform over here. But, by the
116. 5:31end of our pull stage,
117. 5:33we have all the input raw files that we
118. 5:37want inside HubMeta.
119. 5:40In the pick stage, this is where we sift
120. 5:43through our articles, and we pick the
121. 5:46ones that match our inclusion exclusion
122. 5:48criteria. We will go through their
123. 5:50titles, abstracts. We will then find the
124. 5:54PDFs. We will screen their PDFs and in
125. 5:57the end we will have our final sample of
126. 6:00includable
127. 6:01articles. In the previous stage, this is
128. 6:04where we extract data from those
129. 6:07included articles that we have a PDF
130. 6:09from
131. 6:10and we will set up
132. 6:13extraction code books and then we will
133. 6:15set up our AI to extract from these PDFs
134. 6:20and make it ready for the next step,
135. 6:22which is analysis. And inside push,
136. 6:25which is the fourth step, this is where
137. 6:28we do multiple things. We can manage the
138. 6:31taxonomy and that is where we look at
139. 6:34our measurements and constructs and how
140. 6:36they are related to each other.
141. 6:38We will go through our analysis. We have
142. 6:41a lot of analysis implemented here or
143. 6:44you can export your data to be processed
144. 6:48in other software like R, NVivo, Stata,
145. 6:52whatever you're using to do your
146. 6:54analysis. And that would pretty much be
147. 6:57the four steps of
148. 7:00a systematic review.
149. 7:02Let's review them one by one.