Exporting Your Data for Analysis | HubMeta Tutorial #18 — Transcript

Full transcript

1. 0:15Welcome back. Now, we are in the push
2. 0:19stage. We talked about taxonomy. Next
3. 0:22thing we want to learn is how to export
4. 0:24our data out of HubMed. We go to push,
5. 0:29export. And what we see in this page is
6. 0:32at the top, we see different ways to
7. 0:36filter our papers. So, for example,
8. 0:39let's say I want to export only the
9. 0:41papers for which I have extracted a
10. 0:44correlation table. So, I will check this
11. 0:47one. That means only 498
12. 0:51papers will be exported.
13. 0:54Or if I can if I want to search, I can
14. 0:57search like, you know, papers from a
15. 1:00specific journal
16. 1:03to be included in this extraction. Or,
17. 1:07you know, the title or something else.
18. 1:10So, you you get the idea. So, first you
19. 1:12filter.
20. 1:13In advanced filters, these are the same
21. 1:16filters that we saw in our library page.
22. 1:19In the advanced one, you can add groups
23. 1:23here. And here it just says, "Okay, go
24. 1:26to
25. 1:27for example that
26. 1:29one, which was
27. 1:32sample characteristics, and if the total
28. 1:34n was, I don't know, like sample size
29. 1:37was more than 200,
30. 1:40just, you know, give me the export for
31. 1:42those."
32. 1:43But let's extract everything for now.
33. 1:47So, let's go with, you know, all the
34. 1:49ones that have a correlation table, 498.
35. 1:53The after we have filtered, we have
36. 1:56three options to export. The one that we
37. 1:59are most interested in is extracted
38. 2:02data. Like remember the data that was
39. 2:04shown in our capture page and library,
40. 2:07like the ones that AI or ourselves
41. 2:09extracted from each paper. If I click on
42. 2:12generate export, it will
43. 2:16take a few minutes. It will read all of
44. 2:19those data from our database and create
45. 2:23one
46. 2:24and create only one CSV file for us.
47. 2:29And what we can do with that is
48. 2:32absolutely everything. We download that,
49. 2:35we take it outside HubMeta. We can take
50. 2:37it to R, to
51. 2:40Stata, wherever we do our analysis. Uh
52. 2:44it's actually not a bad idea to do all
53. 2:47of your analysis outside HubMeta.
54. 2:50Um
55. 2:51Uh you can go to our analysis page and
56. 2:53see our example code there. But you
57. 2:56know, with R you can do a whole lot of
58. 2:59things that you want to do
59. 3:01on your
60. 3:03data over there. Here we go. The data is
61. 3:08ready. It says 13,000 rows, 86 megabytes
62. 3:13of data. I can download it. Let's
63. 3:15quickly take a look at what our data
64. 3:19looks like.
65. 3:20It's a CSV file. It's a very big one, of
66. 3:25course. 13,000 lines and many, many
67. 3:27columns.
68. 3:29Each line is an effect size in the CSV
69. 3:32file. You have in the columns many
70. 3:35different things, including the journal,
71. 3:37year, sample size, the correlation, the
72. 3:41mean and the standard deviation of the
73. 3:44two sides of the effect size, and all
74. 3:48the data from your templates, like
75. 3:51publication type. Like remember all of
76. 3:54these were things that we
77. 3:57extracted from our papers, like journal
78. 4:01type, uh country,
79. 4:04um
80. 4:05industry, you name it. Like a whole lot
81. 4:08of different things
82. 4:10uh that we extracted. All of them are
83. 4:13available here. In my case, it's
84. 4:16you know, close to a more than hundreds
85. 4:19of um columns.
86. 4:22And
87. 4:23you might or might not use it in in your
88. 4:27analysis, but it's just pure text and
89. 4:30data. You can do anything you want with
90. 4:33this.
91. 4:33The other way to um
92. 4:36to extract information is if you just
93. 4:39want a list of your papers, you can just
94. 4:44you know, use this. This will just give
95. 4:47a list of all the final sample of your
96. 4:49papers, the title, the author, you know,
97. 4:53complete bibliography. This is useful
98. 4:55when you are, you know, creating a final
99. 4:58table to report the papers that were
100. 5:00included in your systematic review.
101. 5:03Finally, with this you can create a zip
102. 5:05file for all the PDFs that were
103. 5:09included. This is, you know, just
104. 5:11basically getting all the PDFs you had
105. 5:13already uploaded on HubMeDa and
106. 5:16downloading them for yourself.
107. 5:18Just mind you, this link that we create
108. 5:21for you to export your PDFs will expire,
109. 5:24I think, in 24 or 48 hours. So, after
110. 5:27you
111. 5:28created your export link,
112. 5:31make sure you come back and download it.
113. 5:33It will take a few minutes to build
114. 5:35that, but
115. 5:36it will be available here for 2 days.
116. 5:39You then take it out of your project.
117. 5:42So, that's how you take your data out of
118. 5:45HubMeDa. You can take out absolutely
119. 5:48everything you have used Hub Meter to
120. 5:51create, and then do it in
121. 5:55and then use it for your analysis and
122. 5:57reporting.
123. 5:58In the final video, I will talk about
124. 6:01our analysis page and what we have
125. 6:03available for you.
126. 6:15>> [music]