Sharing presentations on YouTube

Recently, I was invited to give a presentation at “US Wahl-Watching 2016”, an event to watch the US presidential elections together with our students jointly organized by the institutes of “English and American Studies” and “Political Science” at the Technische Universität Dresden (see here). Given that it was a raving success, I have posted some pictures on Twitter live (see, for example, here and here). My presentation centered around the “agency-structure debate” in international relations (IR) and asked to which extent Mr. Trump will be able to shape world politics by identifying various factors that restrict his scope of action.

While I only had a couple of minutes to talk about each theory and some major theories were omitted altogether (e.g. critical theory), in hindsight I could have been more explicit about the different “levels of analysis” echoed in the agency-structure debate (cf. Steve Saideman’s post here on Duck of Minerva on essentially the same topic published shortly after my presentation). Anyway, since Mr. Trump has won the elections and my students have become quite apprehensive of what this could mean, I have uploaded the presentation on YouTube where it has been watched over 100 times within a week of it being published.

Another motivating factor for me to go through the whole process of putting something online on YouTube was that I was curious to see how burdensome it would be. Overall, I was rather surprised how easy it is. I used PowerPoint 2016 to prepare, record and publish the presentation. Each slide could be recorded individually and as many times as was necessary. While recording, I could have highlighted text (even though I did not use that feature in this presentation) and could easily access my notes on screen without printing them out first. Strangely enough, this only worked on my home computer even though my work computer also runs Office 2016. If I noticed a typo after recording a slide, I could edit the error without losing the audio recording. Better still, I could easily add or replace slides. One simply adds a slide and then records the missing audio track for that slide only. When saving the presentation in video format, PowerPoint processes slides one by one. This means that it would be easy to do something similar for a lecture in a flipped-classroom scenario. When I want to update individual slides the following year this could be easily done.

To boost audio quality, I got a USB condenser microphone for 60 euros (the Auna MIC-900B USB) which was definitely worth the investment. Of course, one does not get studio quality for that kind of money but it is much better than some other solutions for under 100 euros that I have tried. At the end, I set up my own YouTube channel, which was easy enough. The only step taking some time was creating a custom banner image (which I did using GIMP). Among the most critical issues for me was whether any of the graphic material used in the presentation was copyrighted. To get around legal issues, I have included a short disclaimer at the start (around minute four). So far, no one has lodged an injunction against me. Making educational videos freely available is an exciting public outreach opportunity that hopefully more scientists will take advantage of in the future. Moreover, I might also prepare short presentations of my published work to reach out to a greater audience, particularly outside academia.

Finally, the presentation gave me the chance to take a closer look at the Correlates of War Project and data visualization in R. The two figures around minute fourteen and twenty-one have been prepared with Hadley Wickham’s unsurpassed “grammar of graphics” package (ggplot2). Obviously, the two figures are nothing to brag about. Eventually, I did not even play around with different customization options (“themes”) to enhance their appearance. Then again, given that I am really just beginning to see how R can benefit my research I was content to create basic graphs displaying selected data from a data set widely used in IR. For the sake of transparency, here the code to reproduce the figures.


ds <- read.csv("./cow/IGO_stateunit_v2.3.csv")
ds[ds==-9] <- NA
ds[ds==-1] <- NA
ds[ds==2] <- 1  # 2 equals associate membership, which I equate with full membership for this purpose
ds[ds==3] <- 1  # 3 equals observer status, which I equate with full membership for this purpose
ds[ds=="Germany West"] <- "Germany"

# Since all forms of (partial) membership have been recoded 1, row sums correspond to the number of IGOs in which a country is involved. 
ds <- filter(ds) %>%
mutate(sumRow = rowSums(.[5:533], na.rm = TRUE))

# Drop the one observation for Germany 1990 appearing twice (after renaming "Germany West" to "Germany"). Oddly enough, both have entries with competing values for the year 1990. I took the one that fit with the general trend.
ds <- ds[-7862, ]

selectedCountries <- c("China", "Germany", "India", "Japan", "Russia", "USA")

g <- ggplot (subset (ds, country %in% selectedCountries), aes (x = year, y = sumRow, colour = country))
g <- g + geom_line (size = 1)
g <- g + coord_cartesian (xlim = c(1945,2005), ylim = c(0:110))
g <- g + labs (x = "Year", y = "Membership IGOs")
g <- g + scale_colour_manual (values = c("orange", "yellow", "green", "violet", "red", "blue"))
ggsave ("./plots/IGO_stateunit_v2.3_6.png")

Leave a Comment