Resul Umit

This is surprisingly hard to find! I ran into a similar problem last year and couldn’t find any straightforward OLS example for an introductory methods module, so I ended up writing up analysis myself using the BES data.

GitHub - saschagobel/legislatoR: Interface to the Comparative Legislators Database Interface to the Comparative Legislators Database. Contribute to saschagobel/legislatoR development by creating an account on GitHub.

Yes, Wikipedia is indeed an excellent resource for data on MPs across many countries, and it has been put to great use in projects such as the legislators dataset and package (github.com/saschagobel/...).

For our purposes, though, it is a bit less helpful when it comes to variables on candidates.

A scanned historical document titled 'Returns Relating to Election Charges', showing a table of candidates' names, expenses (in pounds, shillings, and pence) for agents, conveyances, advertising, and other costs, along with average cost per vote polled and number of votes received. The print is in small serif font, with some text faint or blurred due to poor scan quality.

Thanks to Chrome’s new OCR and Tabula, and a bit of elbow grease, we can now turn scanned return charges and campaign spending records like this into tidy data.

I’m beyond happy.

How Do Voters Form Expectations about the Parties' Chances of Winning the Election?* Objective. This article examines the factors that form voters' perceptions of the parties' chances of winning at both the national and the local levels. Method. We make use of the 1988 Canadian Ele....

Indeed — voters seem to rely on different sources of information at different levels: this paper suggests that past elections have a particularly strong effect on expectations for constituency results, while opinion polls matter primarily for national-level expectations.

doi.org/10.1111/j.15...

Join "Bluesky Data: Collection, Cleaning, and Analysis" with @resulumit.bsky.social on Aug. 26-27! Learn to access the #Bluesky API, clean social media data, and apply text & network analysis using #Rstats in this hands-on course.

Thanks so much, Catherine! These are really helpful indeed.

I've spent most of the day chasing citations for the claim that waiting for a service negatively affects evaluations of the provider.

Dozens and dozens of papers repeat it and cite a bunch of others --- but when you read the originals, there's no such evidence there at all.

I don’t have a specific example, but I would start by looking at papers citing the "Boix–Miller–Rosato dichotomous coding of democracy" --- either the original article (doi.org/10.1177/0010...) or the updated dataset (doi.org/10.7910/DVN/...).

Unless I have misunderstood the question, of course.

APSR has recently reintroduced its Research Notes track:

bsky.app/profile/apsr...

Thank you!

For elections since 1885, we plan to include a measure of constituency area (un)change, using data from @parlconst.bsky.social.

As for notional results, I'm unsure whether area change alone is informative enough to estimate past outcomes under new boundaries. Or is it? 🤔

A screenshot of R console output showing the names of variables in the dataset. It includes the following 10 variables: "election", "year", "date_start", "date_end", "constituency", "party", "candidate", "unopposed", "votes", and "turnout".

Currently, there are ten variables in the dataset.

What other candidate-level variables could we add, and from which sources?

Suggestions welcome!

I've received funding to improve my UK General Elections dataset (1832–2024)!

We will validate existing variables and add new ones — e.g. registered voters, campaign spending, candidate gender, IDs for matching with other datasets.

Current version:

doi.org/10.7910/DVN/...

I've updated the Political Science Journals app to reflect this.

Curious to see which journals publish which formats, and at what lengths? Explore here:

🔗 resulumit.shinyapps.io/psjournals/

Wikipedia has the exact date of birth for 372 MPs and approximate dates for 83 more. I’ve compiled this into a file and will email it to you in case it’s helpful, even if incomplete.

Line charts showing support for joining vs. remaining outside the EU in hypothetical referendums for non-EU countries (Iceland, Montenegro, Norway, Serbia, Switzerland). Green lines (join) rise in some countries, while red lines (remain outside) decline. Data from ESS.

Non-EU member states show some of the most striking shifts in public opinion on EU membership.

In a hypothetical referendum on whether their country should join or remain outside the EU, support for joining has noticeably increased over time in most surveyed countries.

A grid of line plots showing support for EU membership ("remain") and leaving the EU ("leave") in hypothetical referendums across 10 EU countries. Green lines indicate support for remaining, consistently higher than red lines for leaving, with green increasing slightly over time in most cases. Data from ESS.

How would people vote if there were a Brexit-like referendum in their country?

Support for EU membership remains strong across the EU and, if anything, has been increasing in many countries since the Brexit referendum.

A screenshot of an interactive web application titled "Political Participation in Europe." The interface allows users to visualise data on political participation. The left panel includes dropdown menus, with "Level of Analysis" set to Europe and "Participation Forms" allowing multiple selections. Two options, "Attended a demonstration" and "Boycotted a product," are selected from the dropdown list. A longitudinal line chart on the right visualises trends for these two forms of participation from 2002 to 2023. The red line represents "Boycotted a product," and the blue line represents "Attended a demonstration." The y-axis ranges from 0% to 75%, and tabs at the top provide options for "Longitudinal Analysis" (currently selected), "Subgroup Analysis," "Correlation Analysis," and "Notes."

A screenshot, showing the "Subgroup Analysis" tab. The analysis is set to "Regions," with four regions selected, focusing on university students or graduates who voted in the last national election in 2023. The visualisation displays bar charts comparing participation across Central and Eastern Europe, Northern Europe, Southern Europe, and Western Europe.

A screenshot of the "Correlation Analysis" tab. The scatter plot visualises the relationship between "Contacted a politician" (x-axis) and "Voted in the last national election" (y-axis) at the European level for the year 2023. A regression line is included, with data points scattered around it, showing a positive correlation. The x-axis ranges from 0% to 20%, and the y-axis ranges from 0% to 100%. Selection options for variables and year are displayed in the left panel.

Another screenshot of the "Longitudinal Analysis" tab, this time the analysis focuses on countries, with 37 countries selected and two participation forms: "Contacted a politician" (red) and "Voted in the last national election" (blue). The visualisation includes small multiple line charts for each country, such as Austria, Belgium, Bulgaria, and Croatia, showing trends over time for the selected participation forms. The y-axis ranges from 0% to 75%.

I’ve updated my Shiny app on political participation with the latest ESS data released yesterday.

I use this app for teaching, to help students explore and analyse patterns of political participation.

resulumit.shinyapps.io/participation/

#rstats

Thank you! You're absolutely right that not everyone needs a visitor visa to travel to Canada for a conference, and one should check.

My post was about those who would need one, like myself.

Are you planning to propose a paper for #APSA2025, and you would need a visa to travel to Canada?

Do check the visa processing times first. As of now, the processing time for a visitor visa application from the UK is 398 days!

www.canada.ca/en/immigrati...

The table lists political science journals that accept data-focused articles. It includes columns for the journal name, h5-index, format name, and word limit for submissions. It's a screenshot from the linked webpage.

Do you have a dataset that the political science community should know and use?

At least 18 journals in the discipline now offer a dedicated article category for sharing datasets. Find out more on my blog:

resulumit.com/blog/polisci...

Table listing political science journals that publish replication studies. Columns include Journal, h5-index, Format Name, and Length Limit. Journals listed are: American Political Science Review (h5-index 82, no length limit), Public Administration (h5-index 47, no length limit), Political Communication (h5-index 46, 30-page limit), Political Science Research and Methods (h5-index 44, no length limit), Research & Politics (h5-index 29, 2,000-word limit), Canadian Journal of Political Science (h5-index 25, 2,000-word limit), and Journal of Experimental Political Science (h5-index 25, 3,000-word limit).

Seven political science journals now offer a dedicated category for replication studies! Each of these journals accepts replications of articles published both within their own journal and in other relevant journals.

resulumit.com/blog/polisci...

A bar chart titled "Has the incumbent extended their term or eligibility to run?" The x-axis represents years, ranging from before 1960 to after 2020. The y-axis shows the count of occurrences, reaching up to about 25. Bars are stacked and colour-coded: yellow represents "No," indicating cases where incumbents did not extend their terms, while dark green represents "Yes," showing instances where incumbents extended their terms or eligibility. Over time, the yellow bars dominate, suggesting more elections without term extension, but green segments are visible, especially towards the later years. At the bottom right, a caption reads "Data: nelda.co | Graph: Resul Umit."

With the chatter yesterday about a possible third Trump term, I looked at data on how often sitting incumbents around the world have extended their term or eligiblity to run.

Cases have become more common since about 2000, though this coincides with a general rise in executive elections globally.

Thank you very much for reporting this.

It seems a last-minute change in code affected that filter without me realising. I’ve now resolved the issue, and the updated version is live.

Screenshot of a web app titled 'Political Science Journals' by Resul Umit, designed to filter data on political science journals. The interface includes several filtering options on the left, such as 'Journal scope' (a text field), 'Publishing since' (slider from 1837 to 2024), 'Publisher' (dropdown menu with selected items), 'Impact Factor' (slider from 0 to 10), 'H5 Index' and 'H5 Median' (both with sliders from 0 to 119 and 0 to 220, respectively). There is also an option to 'Include journals with unknown metrics', a dropdown for 'Article type' (with 8 items selected), and a field for 'Length limits' with a unit selection of 'Words'.

Looking for the right political science journal for your next paper? 📄✨

I've updated my app with new data and functionality to help you navigate over 300 journals by scope, article types, and more.

resulumit.shinyapps.io/psjournals/

Author Guidelines International Affairs is a leading anonymous peer-reviewed journal publishing academically rigorous, policy relevant articles covering the entire field of Inter

A great initiative from International Affairs:

Board members mentor selected early-career, under-represented scholars in International Relations, supporting them in strengthening their manuscripts and navigating the publishing process.

academic.oup.com/ia/pages/Aut...

It's something provided by Google Scholar, indicating how often a journal's articles are cited over the last five years: an h5-index of 20 means that 20 articles published within that period have each been cited at least 20 times.

I'm glad to hear it is useful.

A screenshot from the linked website, showing a table listing political science journals that publish research notes, displaying four columns: Journal, Format Name, Length, and h5-index.

I have updated the list of political science journals that publish research notes.

resulumit.com/blog/polisci...

A screenshot of the dataset repository on Harvard Dataverse. It shows that the dataset is titled "UK House of Commons Election Results at Candidate Level". It contains candidate-level election results for the UK House of Commons between 1832 and 2024, with 89,545 observations and 10 variables. The dataset is under a public domain license.

The results from 49 elections to the UK House of Commons in one dataset, now including the 2024 General Election.

89,000+ observations of UK electoral history.

doi.org/10.7910/DVN/...

#polisky #politicalscience