Video as data & documentation
Rick O. Gilmore
2019-09-18 12:00:38
Preliminaries
Overview
- Video as data
- Video as documentation
- Sharing ethically & securely with Databrary
- Accelerating discovery
Video as data & documentation
Video as data
If a picture is worth 1,000 words…
A video is worth…
…a paper in Science
Video…
- Captures (& preserves)
- Shows (& helps tell…)
- Expands the scope of inquiry
- Provides unparalleled opportunities for reuse
Video as documentation
Even audio
Shared video & audio recordings
- Make research more transparent
- Bolster the reproducibility of procedures
- Accelerate the adoption of new research techniques
- Strengthen findings
- Maximize the return on public investments in research
Sharing video ethically & openly
Video must be protected
- Faces, voices
- Home interiors
- Behaviors that may embarrass participants
Open sharing advances discovery
- Citation expected, but
- No requirement for co-authorship
- No pre-approval of research questions
Requires…
- researchers secure participant permission to share
- IRB/ethics board approval to share
- institutional approval for access
Permission to share
Restricting access
- Researchers at institutions
- Formalized by an institutional agreement
- Protects participants, researchers, & institutions
Standardized levels of sharing
Accelerating discovery
Need
- Annotation tools
- More (bigger, denser, diverse) data
- Platforms for discovery
- \(n=900\) infant/mother dyads, \(n=300\) 12-, 18-, and 24-month-olds
- Demographics, health status, media use, temperament
- Modified MB-CDI (interview, video-recorded, English/Spanish item-level equivalents)
- Open protocol, video documentation
Repositories as Platforms for discovery
vol_8 <- databraryapi::download_session_csv(vol_id = 8)
xtabs(formula = ~ participant.gender + participant.race, data = vol_8) %>%
knitr::kable(.)| Asian | Black or African American | Unknown or not reported | White | ||
|---|---|---|---|---|---|
| 7 | 0 | 0 | 0 | 0 | |
| Female | 0 | 112 | 341 | 4 | 209 |
| Male | 0 | 111 | 410 | 2 | 155 |
race_gender <- vol_8 %>%
filter(participant.gender %in% c('Male', 'Female')) %>%
ggplot() +
aes(x = participant.race, fill = participant.race) +
facet_grid(. ~ participant.gender) +
geom_bar(stat="count") +
coord_flip() +
theme(axis.text.x = element_text(angle = 90,
hjust = 1, size = 16),
axis.text.y = element_text(size = 16),
axis.title.y = element_blank(),
axis.title = element_text(size = 16),
strip.text = element_text(size = 16),
legend.position = "none")
## [1] TRUE| name | asset_type | asset_id |
|---|---|---|
| childcare | Comma-separated values | 159177 |
| locomotion | Comma-separated values | 117092 |
| child-birth | Comma-separated values | 116791 |
| family | Comma-separated values | 116790 |
| sleep | Comma-separated values | 116789 |
| language-exposure | Comma-separated values | 116787 |
lang_exp <- databraryapi::read_csv_data_as_df(session_id = play_session_id,
asset_id = 116787)
lang_exp_plot <- lang_exp %>%
ggplot(.) +
aes(x = language, fill = language) +
facet_grid(exposure_context ~ .) +
geom_histogram(stat='count') +
theme(axis.text = element_text(size = 15),
strip.text = element_text(size = 15),
axis.title = element_blank(),
legend.position = "none")
Video is just one type of time series
Next steps…
Busting some myths
- Data X can’t be shared…
- Open sharing means public
- My data are mine
Each one, reach one
“SRCD regards scientific integrity, transparency, and openness as essential for the conduct of research and its application to practice and policy.”
“These values apply to research conduct, to the teaching of scientific methods, and to the translation of science into practice and policy.”
“SRCD regards it as an essential responsibility of the producers and consumers of child development research to act ethically, conscientiously, and with integrity…”
“…and to strive for transparency and openness in research practice, publication, and dissemination.”
Investing in the future
- …of robust & reproducible social science
A modest proposition: The 15% solution
Federal funding agencies should set aside…
- 5% of the research budget for data repository operational support
- 5% of the research budget for data curation activities
- 5% of the research budget for secondary analyses of shared data
Thank you
rogilmore@psu.edu https://gilmore-lab.github.io https://gilmore-lab.github.io/2019-08-28-apsa/ @rogilmore
Materials
This talk was produced on 2019-09-18 in RStudio version 1.2.1335 using R Markdown and the reveal.JS framework. The code and materials used to generate the slides may be found at https://github.com/gilmore-lab/2019-08-28-apsa/. Information about the R Session is as follows:
## R version 3.5.3 (2019-03-11)
## Platform: x86_64-apple-darwin15.6.0 (64-bit)
## Running under: macOS Mojave 10.14.6
##
## Matrix products: default
## BLAS: /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib
## LAPACK: /Library/Frameworks/R.framework/Versions/3.5/Resources/lib/libRlapack.dylib
##
## locale:
## [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
##
## attached base packages:
## [1] stats graphics grDevices utils datasets methods base
##
## other attached packages:
## [1] qrcode_0.1.1 forcats_0.4.0
## [3] stringr_1.4.0 dplyr_0.8.1
## [5] purrr_0.3.2 readr_1.3.1
## [7] tidyr_0.8.3 tibble_2.1.3
## [9] ggplot2_3.2.0 tidyverse_1.2.1
## [11] databraryapi_0.1.6.9001
##
## loaded via a namespace (and not attached):
## [1] tinytex_0.15 revealjs_0.9 tidyselect_0.2.5
## [4] xfun_0.8 reshape2_1.4.3 haven_2.1.0
## [7] lattice_0.20-38 colorspace_1.4-1 generics_0.0.2
## [10] htmltools_0.3.6 base64enc_0.1-3 yaml_2.2.0
## [13] rlang_0.4.0 R.oo_1.22.0 pillar_1.4.1
## [16] glue_1.3.1 withr_2.1.2 R.utils_2.9.0
## [19] modelr_0.1.4 readxl_1.3.1 plyr_1.8.4
## [22] munsell_0.5.0 gtable_0.3.0 cellranger_1.1.0
## [25] R.methodsS3_1.7.1 rvest_0.3.4 codetools_0.2-16
## [28] evaluate_0.14 labeling_0.3 knitr_1.23
## [31] curl_3.3 highr_0.8 broom_0.5.2
## [34] Rcpp_1.0.1 scales_1.0.0 backports_1.1.4
## [37] webshot_0.5.1 jsonlite_1.6 mime_0.7
## [40] hms_0.4.2 packrat_0.5.0 digest_0.6.19
## [43] stringi_1.4.3 keyring_1.1.0 grid_3.5.3
## [46] cli_1.1.0 tools_3.5.3 magrittr_1.5
## [49] lazyeval_0.2.2 crayon_1.3.4 pkgconfig_2.0.2
## [52] xml2_1.2.0 lubridate_1.7.4 assertthat_0.2.1
## [55] rmarkdown_1.13 httr_1.4.0 rstudioapi_0.10
## [58] R6_2.4.0 nlme_3.1-140 compiler_3.5.3