Lightnews — Scholar-powered news

Julia M. Rohrer

@dingdingpeng.the100.ci

10K followers 2.4K following 8.3K posts

Personality psych & causal inference @UniLeipzig. I like all things science, beer, & puns. Even better when combined! Part of http://the100.ci, http://openscience-leipzig.org

the100.ci,

Posts Media Videos Starter Packs

Pinned

Julia M. Rohrer @dingdingpeng.the100.ci · Aug 25

Ever stared at a table of regression coefficients & wondered what you're doing with your life?

Very excited to share this gentle introduction to another way of making sense of statistical models (w @vincentab.bsky.social)
Preprint: doi.org/10.31234/osf...
Website: j-rohrer.github.io/marginal-psy...

Models as Prediction Machines: How to Convert Confusing Coefficients into Clear Quantities

Abstract
Psychological researchers usually make sense of regression models by interpreting coefficient estimates directly. This works well enough for simple linear models, but is more challenging for more complex models with, for example, categorical variables, interactions, non-linearities, and hierarchical structures. Here, we introduce an alternative approach to making sense of statistical models. The central idea is to abstract away from the mechanics of estimation, and to treat models as “counterfactual prediction machines,” which are subsequently queried to estimate quantities and conduct tests that matter substantively. This workflow is model-agnostic; it can be applied in a consistent fashion to draw causal or descriptive inference from a wide range of models. We illustrate how to implement this workflow with the marginaleffects package, which supports over 100 different classes of models in R and Python, and present two worked examples. These examples show how the workflow can be applied across designs (e.g., observational study, randomized experiment) to answer different research questions (e.g., associations, causal effects, effect heterogeneity) while facing various challenges (e.g., controlling for confounders in a flexible manner, modelling ordinal outcomes, and interpreting non-linear models).

Figure illustrating model predictions. On the X-axis the predictor, annual gross income in Euro. On the Y-axis the outcome, predicted life satisfaction. A solid line marks the curve of predictions on which individual data points are marked as model-implied outcomes at incomes of interest. Comparing two such predictions gives us a comparison. We can also fit a tangent to the line of predictions, which illustrates the slope at any given point of the curve.

A figure illustrating various ways to include age as a predictor in a model. On the x-axis age (predictor), on the y-axis the outcome (model-implied importance of friends, including confidence intervals).

Illustrated are
1. age as a categorical predictor, resultings in the predictions bouncing around a lot with wide confidence intervals
2. age as a linear predictor, which forces a straight line through the data points that has a very tight confidence band and
3. age splines, which lies somewhere in between as it smoothly follows the data but has more uncertainty than the straight line.

49 280 940

Reposted by Julia M. Rohrer

Naomi Saphra @nsaphra.bsky.social · 8h

It is wild that anyone out there is advocating for MORE human labor in image moderation. a friend of mine used to be a 4chan janitor. they’re dead now, but they spent the rest of their life unable to look at a child without feeling sick.

2 11 40

Julia M. Rohrer @dingdingpeng.the100.ci · 5h

If we add “bath tub” to washing machine and dishwasher, then I guess it covers virtually anything, including kids 😂

Reposted by Julia M. Rohrer

Ian Hussey @ianhussey.mmmdata.io · 17h

Please someone help me figure out the rest of this joke

4 4 33

Julia M. Rohrer @dingdingpeng.the100.ci · 18h

uvu kvantitativ

Julia M. Rohrer @dingdingpeng.the100.ci · 18h

Prank a German by setting their spell checker to Dutch. Hilarity ensues.

Julia M. Rohrer @dingdingpeng.the100.ci · 19h

Excuse me?

Text snippet: "quantitative approaches are..."
With word auto-suggesting the alternative spelling "kwantitatieve, kwantitatiever"

3 3 27

Julia M. Rohrer @dingdingpeng.the100.ci · 19h

I did not *but* I copied a snippet from an (English) document written by somebody in Belgium, so I can see how that happened. My Word is really bad about languages, like it will insist I'm trying to write something in German even after I repeatedly marked the text as English.

1 1

Julia M. Rohrer @dingdingpeng.the100.ci · 19h

Happy to correct my statement to "read more outside of their field, but also the classics in their field"

1 3

Julia M. Rohrer @dingdingpeng.the100.ci · 19h

They are 😂 I just realized that.

Julia M. Rohrer @dingdingpeng.the100.ci · 19h

Dutch, apparently: www.linguee.com/english-dutc...

quantitative - Dutch translation – Linguee

Many translated example sentences containing "quantitative" – Dutch-English dictionary and search engine for Dutch translations.

www.linguee.com

Julia M. Rohrer @dingdingpeng.the100.ci · 19h

I read that book and liked it a lot! That's part of why the other hypothesis seems so bizarre to me 🙈

1 1

Julia M. Rohrer @dingdingpeng.the100.ci · 19h

omg is this Dutch? I thought the spell checker was having a mental breakdown

2 21

Julia M. Rohrer @dingdingpeng.the100.ci · 19h

Excuse me?

8 7 59

Julia M. Rohrer @dingdingpeng.the100.ci · 23h

henlo fren

VGA adapter with the letters uwu superimposed

Julia M. Rohrer @dingdingpeng.the100.ci · 23h

Funnily enough, apart from the evolutionary angle, this feels like it's moving towards the type of self-care advise usually found in women's magazines 💅🏼 "Women, treat yourself" says Jordan Peterson.

1 5

Julia M. Rohrer @dingdingpeng.the100.ci · 1d

which happens to be my catchphrase whenever I enter a classroom

2 2

Julia M. Rohrer @dingdingpeng.the100.ci · 1d

"Who wore it better", Seraphim vs Cherubim

Julia M. Rohrer @dingdingpeng.the100.ci · 1d

Indeed, I think the world couldn't do without all that online-sample-imagined-intervention-self-report-outcome research that is urgently needed because of *gestures vaguely* important implications for contemporary issues.

1 1 5

Julia M. Rohrer @dingdingpeng.the100.ci · 1d

I really strongly feel that some fields of research would profit if researchers stopped collecting online data for some time and instead maybe just read a bit outside of their field.

The ‘harm hypothesis’ strikes me as being deeply rooted in contemporary WEIRD values rather than being the result of a specific ‘evolved’ or ‘innate’ instinct or psychological mechanism. And indeed the literature cited to support it seems to suggest this.

Costello & Acerbi cite 5 papers in the paragraph above to support the model:

Stewart-Williams et al., 2024: the sample here consists of Prolific users mostly in the UK.

FeldmanHall et al., 2016: the samples were MTurk users in the US and volunteers in the UK.

Curry et al., 2004: the sample is convicted offenders in Texas in 1991.

Graso et al., 2023: US MTurk users again.

Graso & Reynolds, 2024: this is a review paper which does make some cross-cultural claims, but when you check the references you can see some important limitations. For example, they write that “Across cultures, women were perceived as less powerful than men but were seen more positively,” and when you check the reference it goes to Glick et al., 2004, which samples from 16 nations. However, when you read that paper they note in the methods that, “Most samples consisted primarily of college students participating for extra credit.”

3 7 39

Julia M. Rohrer @dingdingpeng.the100.ci · 1d

Evolutionary psychologists have claimed that "humans evolved heightened sensitivity to harm directed at women." This excellent (albeit at times gruesome) post makes the point that the ethnographic evidence does not really support this narrative.

Did humans evolve to 'protect' women?

Ethnography complicates a convenient narrative

traditionsofconflict.substack.com

5 14 87

Julia M. Rohrer @dingdingpeng.the100.ci · 1d

Just put on a dress with a slightly disorienting pattern involving what appears to be eyes and my husband complimented me for looking like a biblically accurate angel 👁️ 👁️ 👁️ 👁️

3 51

Julia M. Rohrer @dingdingpeng.the100.ci · 1d

I see how that’s important context for a child of a certain age 👧🏻

Julia M. Rohrer @dingdingpeng.the100.ci · 1d

It’s like time travel, contained in a single device.

Julia M. Rohrer @dingdingpeng.the100.ci · 1d

Precisely!!! Sometimes I forget about my own rule and ask my husband (“can this go into the dishwasher?”) to which he respond “let’s find out”

Julia M. Rohrer @dingdingpeng.the100.ci · 1d

Even the paw patrol movie is too scary for her lolsob

2 1

Reposted by Julia M. Rohrer

John P de New 🇦🇺🇨🇦🇩🇪🇪🇺 @denewjohn.bsky.social · 1d

You're never gonna iron that. Just buy clothes that don't need to be ironed. Don't lie to yourself.

3 4 21