Analyzing Your Data in HubMeta | HubMeta Tutorial #19 — Transcript

Full transcript

1. 0:15Welcome back to our final video in this
2. 0:18series.
3. 0:20I want to talk to you about our analysis
4. 0:22page. What this page is something we
5. 0:25have built based on request from our
6. 0:28community of users. Many times they
7. 0:31wanted us to share with them the codes
8. 0:33that we use for our own analysis. And we
9. 0:37decided to actually include this as a
10. 0:40clean nice way of sharing our codes with
11. 0:44you. This is an under construction page.
12. 0:48We will keep adding more and more of our
13. 0:51methods and codes here. So far we have
14. 0:55two example codes here. But
15. 0:59I will explain here how you can use
16. 1:01this.
17. 1:02First of all, let's say you want to run
18. 1:04any custom analysis inside this. What
19. 1:08this page is like an R studio page. If
20. 1:12you have worked with R, if you have
21. 1:14coded with R, you know that you have one
22. 1:17place to select your environment. Those
23. 1:20are the files that you include in your
24. 1:23project or need to export.
25. 1:25You have a place to write your actual
26. 1:28code. This is the console view where the
27. 1:32output of your analysis is shown. And
28. 1:35then here, write your code and then
29. 1:38click on play. And we will the code will
30. 1:41run for you. It has certain formatting.
31. 1:46Usual R code will work very nice here.
32. 1:49You can get it from any AI or just write
33. 1:54your own AI code or write your own R
34. 1:57code. Very soon we will have this AI
35. 2:00assist button working too, so that you
36. 2:03can use this to write your AI code. But
37. 2:06let me just give you an example from
38. 2:09one of our analysis. What these analysis
39. 2:12down here are are basically our custom
40. 2:14analysis page, but we have an actual
41. 2:18sample code here. But for each of them
42. 2:21we have our actual sample code in it. In
43. 2:24this one for example, for meta-analytic
44. 2:27studies one of the things that we do is
45. 2:29to calculate composite scores for cases
46. 2:33where there is one construct and
47. 2:35multiple measurements that are children
48. 2:38of the same parent within one paper, and
49. 2:41we want to calculate the composite
50. 2:43effect. For that, we need to include a
51. 2:46few files including an export of our
52. 2:49taxonomy and a recent export of our data
53. 2:54file. So, if we check these two, this
54. 2:58analysis is now ready to run. So, if I
55. 3:01click on play, this will send all of
56. 3:04this data and our code to the R server
57. 3:09we have running on Hub Meta. It will
58. 3:12process this and you can see this shows
59. 3:15the pipeline that is running on these
60. 3:19496 papers to generate the data with
61. 3:25composite scores in it.
62. 3:28Let's give it a minute to finish its
63. 3:30job.
64. 3:35All right. Now we can see that the job
65. 3:37is finished and it created the output
66. 3:42files. If we see if we look at this
67. 3:45console and down here you see the actual
68. 3:49files that were generated through this
69. 3:52and this one will be the one that I take
70. 3:55out. So, it really depends on the R code
71. 3:58that you have and the type of analysis
72. 4:00you want, but this essentially gives you
73. 4:03a complete platform to do analysis.
74. 4:06We are constantly developing new
75. 4:09techniques and new ways of doing things
76. 4:11like in this one, it does meta-analytic
77. 4:15correlation tables. We have developed
78. 4:18new ways of data cleaning new ways of
79. 4:21taxonomy, many different effect size
80. 4:23tools. And this part which is on
81. 4:26qualitative analysis, this is something
82. 4:28that we have assembled a fantastic team
83. 4:32who are qualitative researchers and and
84. 4:35are helping us to build the qualitative
85. 4:38analysis version for Hub Meta. Well, we
86. 4:41hope that in the next 3 to 6 months, we
87. 4:43will see an extension of Hub Meta into
88. 4:47qualitative systematic reviews. Thank
89. 4:49you for being with us in this new series
90. 4:53of videos from Hub Meta. Please share
91. 4:57with us if you have any questions, if
92. 5:00you found an issue on Hub Meta, or if
93. 5:03you had any suggestions for things that
94. 5:05we can do better inside the platform to
95. 5:08make it work better
96. 5:09>> [music]
97. 5:10>> for our open science mission.