I’m excited to share that I’ve accepted a position as an assistant professor at
@UChicagoCS
and
@DSI_UChicago
starting in 2022! It’s a dream come true for me. I’m looking forward to joining my amazing colleagues and building my lab, focused on interfaces for data cognition!
I defended my dissertation yesterday! It feels strange to be just about finished with my PhD. Huge thanks to the many people who’ve been a part of my grad school journey, especially my wonderful advisor
@JessicaHullman
, mentors
@mjskay
@jeffrey_heer
@amyjko
, and peers!
Does it bother anyone else that, after working on our research for years, we are expected to transfer copyright to publishers or pay them thousands of dollars to make our work open access? Why do we need publishers? Why do they get to own the fruits of our labor?
I’m growing annoyed at pressure to study, use, and brand everything AI. It’s not even that I’m disinterested in AI, but the concomitant push towards intellectual and cultural conformity in academia at this moment is suffocating.
My feed is full of people who went to CHI last week and tested positive over the weekend. Maybe it was too soon to gather in person by the thousands? Seeing this has pretty much killed whatever FOMO I was feeling last week about missing the conference. I hope people recover soon.
Big news! My paper with
@JessicaHullman
&
@mjskay
on “Visual Reasoning Strategies for Effect Size Judgments and Decisions” won Best Paper at IEEE InfoVis 2020!
See my original thread about the paper for details.
We’re excited to share our preprint, “Visual Reasoning Strategies and Satisficing: How Uncertainty Visualization Design Impacts Effect Size Judgments and Decisions”, with
@JessicaHullman
&
@mjskay
Details in the thread! (1/n)
There’s a trend of academics equating being critical with being mean spirited. As someone whose upbringing cast critique/complaint as an inherently good thing (striving for better, reconciling discomfort/ambiguity), I cannot help feeling somewhat maligned by this rhetoric.
@SteveWechslerPT
@IanArawjo
@AcademicChatter
I’m sorry but is a guilt trip the best response here? People are exhausted and overworked. Service work is voluntary, unless you’re a prof in which case it’s still your prerogative what to say yes to. There are ways of fixing peer review without blaming reviewers for what’s wrong
We’re excited to share our preprint, “Visual Reasoning Strategies and Satisficing: How Uncertainty Visualization Design Impacts Effect Size Judgments and Decisions”, with
@JessicaHullman
&
@mjskay
Details in the thread! (1/n)
We are seeking savvy data analysts to participate in a user study on new visualization tools for exploratory data analysis!
Do you work with data regularly? Are you interested in GUI tools for data exploration? We want to hear from you!
1/
I’m recruiting PhD students this cycle
@UChicagoCS
@DSI_UChicago
. If you’re excited about visualization, if you have burning questions about how people think with data, if you want to create statistical software, apply to work with me!
I am recruiting PhD students and postdocs this year! If you’re excited about building and evaluating data visualization and analysis software, please consider applying to join my lab
@UChicagoCS
@DSI_UChicago
If I worked at Google, I would be looking for a job with a different company. How many more unscrupulous firings before Google is starved of talent and contracts?
Maybe I’m being naive in thinking they can be held accountable, but I hope I’m not.
Submitting to CHI has become an unending list of chores that goes on for months. It should be a paper and a talk, that’s it. ACM is not entitled to this much of people’s time.
Woman thought of the day: unless you time the birth of a child in May exactly, you will never be able to be pregnant AND submit and get a paper accepted at
@acm_chi
if you also want actual maternity leave. The things you have to do post-submission extend from September-April.
As an American citizen and a CS prof, I could not be more opposed to visa nonsense like this. It is unfair. It makes us look weak and petty on the world stage. If we continue like this, the US will alienate talented people, and it will be our loss/failure when they go elsewhere.
Can we talk about how inconvenient it has gotten to get a visa for Chinese students? It’s been 3 weeks since mine was “refused”.
Is there any reason me, in my 2nd yr Stanford PhD, must wait for up to 180 days to be considered safe to be let in the US?
I’m excited to start an internship
@MSFTResearch
next week! I’ll be working with Rich Caruana,
@jennwvaughan
, and the excellent team they’ve assembled on visualization tools for ML interpretability. 📊📈
0 for 2 on co-authored CHI submissions this year. You cannot win them all. Setbacks and resubmissions are completely normal in research. If you’re a grad student feeling bad about rejection, chin up and keep on pushing toward your goals!
Am I the only who’s not enthusiastic about moving to Mastodon? I don’t want another social media site in my life tbh (why I don’t have Instagram, TikTok, etc.) If the community of scientists I follow here dissolves, that might just be the end of this form of communication for me.
There’s a difference between being a jerk and being critical. I think we may be straying too far from “don’t be a jerk” and too close to “hold your tongue about issues to keep others comfortable”.
I'm excited to share our study on HOPs at VIS 2018 w/
@JessicaHullman
@mjskay
&
@_fmnguyen
.
Check out the paper:
Also see our supplemental materials for analysis scripts, anonymized data, code, and links to our preregistrations:
For some, being critical is a way of knowing/being in the world. I cannot help but think we lose important perspective when we are so quick to shut that down, transforming a scholarly discussion into a judgment of individual character.
I’m pleased to announce our
#CHI2023
paper with Sarah Lee, TJ Goan,
@stats_tipton
, and
@JessicaHullman
: “MetaExplorer: Facilitating Reasoning with Epistemic Uncertainty in Meta-Analysis” 🧵
CS researchers: How do you come up with ways to frame your research “vision”? Does it come to you in a dream like the structure of the benzene ring? Do you have to climb a mountain and meditate? What am I missing here?
This article was the argument that finally pushed me over the edge to cancel my travel plans next week, this and seeing rising case numbers. Although I had already been leaning towards canceling, it turns out hope is a more powerful motivator than a desire to do the right thing.
Did you try and fail to to convince your elderly relatives to avoid holiday gatherings? I have a suggestion. Propose waiting until March. Things are grim now so it's a horrid time to get sick but the US could experience a dramatic turnaround. Hope is real.
I canceled half of my holiday plans to err on the side of caution over Omicron… Feels like crap. The FOMO is especially rough the second year in a row, not helped by my elderly parents insisting that this is an overreaction on my part.
I’m excited to announce my IEEE VIS 2023 paper “EVM: Incorporating Model Checking into
Exploratory Visual Analysis” with Ziyang Guo, Emily Qiao,
@jeffrey_heer
,
@JessicaHullman
@SchrodingrsBrat
I think you’re a little off on this one. Maybe some people look to science as a source of absolute truth, but many of us understand it is a process in motion. For many atheists and agnostics, the whole point is a doctrine of skepticism and a rejection of absolute faith.
#chi23
featured some exceptional visualization research! Probably one of the better conferences I’ve attended so far in my career. Although there were a small-ish number of us, I was reminded how wonderful the visualization community is and how lucky I am to be a part of it!
Huge pet peeve in papers: "To our knowledge, we are one of the first to study this [obviously important] question." Why do authors say this? It adds no value, but it betrays hubris. The authors think it's more likely that they are original than that they've overlooked prior work.
Between organizing conferences and peer review, it seems to me that researchers already do all of the necessary work on a volunteer basis. Why don’t we just organize among ourselves and yeet the publishers? Fuck them for taking and commodifying our intellectual property.
Being active in the visualization research community for 5 years, I see the same ideas recycled year after year but presented as if they are new. Can we just drop the pretense of novelty and accept that we are “puzzle solving” (Kuhn) more than we are constantly innovating?
@amyjko
I am in such a state of time scarcity that I get frustrated when I have to take time to sleep or feed myself… I imagine this is worse with the workload of a professor. So peer review is falling a little low on my list of priorities, as I imagine it is for many folks.
@bmwiernik
I’m no fan of the publishing system… but I think submitting papers while refusing to review for the same venue is an abuse of the generosity of peers who do volunteer to review. I see a difference between declining vs refusing to review.
I find it odd when HCI reviewers request disclaimers about generalizability for qualitative work. Like sure, we can put that in limitations, but it kinda goes without saying and maybe sends a misleading signal to the reader. It’s a methodological red herring.
If I wanted to research AI, I would have done a PhD on it, and I probably would have fucked off to industry in order to afford a house. But I chose a different life, one based on the premise of academic freedom. Let me at least define my own research interests!
That feeling when you’re so tired of a project that if the paper doesn’t get into CHI, you might just yeet it into the sun because the opportunity cost of continuing to work on it is too much 🙃
@Philonous
I don’t think so.
1. Feminist scholarship is inherently _critical_ of dominant power structures.
2. I think it’s unfair/unreasonable/factually incorrect to paint women scholars broadly as somehow more emotionally fragile in the face of critique.
I’m recruiting PhD students this cycle
@UChicagoCS
@DSI_UChicago
. If you’re excited about visualization, if you have burning questions about how people think with data, if you want to create statistical software, apply to work with me!
I would argue that boxplots are not so great. Has anyone ever used a boxplot to arrive at an insight about their data that could not have been supported as well or better without reducing a distribution to summary stats?
It is asinine and cruel that we (the US) try so hard to deny entry, freedom of movement, and long-term residency to international students. These folks are hard-working, contributing members of our society. Their presence here enriches our communities!
He was told that he did not “prove” he will go back to his country after his education is done (section 214b). Now you might say how does one even prove that? The state department says you need to show connections to your home country like having family members or a job or …
We are delighted to announce the winners of this year’s Sloan Research Fellowship! These outstanding researchers are shining examples of innovation and impact—and we are thrilled to support them. Meet the winners here: 🎉
#SloanFellow
#STEM
#ScienceTwitter
I’m looking for participants to evaluate experimental software for meta-analysis. If you have experience with scientific review and/or meta-analysis, please test out our tool. w/
@JessicaHullman
@stats_tipton
Please DM me or email me at kalea
@uw
.edu if you’re interested.
Feeling cognitive dissonance about generative AI. On the one hand, I see a lot of people I respect excited about it. On the other hand, I cannot shake the feeling that we are approaching an event horizon of meaninglessness. What is the value of creative work not done by people?
It was a joy to watch
@_mcnutt_
successfully defend his dissertation this afternoon.
@UChicagoCS
was fortunate to have you as a doctoral student. Huge congrats to Andrew!
UChicago is hiring faculty and staff in CS and Data Science! We have lots of openings this year for folks with different strengths and levels of experience.
I just attended a panel on faculty job negotiations, and everyone avoided answering the most upvoted questions: “What should you expect in dollars for signing, start-up, and salary?”
Why is it so hard to get up-to-date information about this in your particular research area?
The issues summarized in this blog post put a finger on how doing ML/data science without understanding stats is a recipe for unreliable analysis. Imo there’s no such thing as “learn to be a data scientist in 10 weeks”, no shortcuts to expertise or mastery.
Imagine paying more than the average person earns in 10,000 lifetimes to demolish one of the largest and most influential communication platforms in human history, seemingly out of reactionary egotism. No one should have this power.
It was a great week at
#ieeevis
! Thanks to all the people who spent time catching up with me and chatting with my students about research. It means so much to be a part of a vibrant and welcoming research community.
This articulates my misgivings about “data mining” pretty clearly. We need data analysis tools that promote interrogating the meaning behind the apparent patterns we detect with visualization and ML. It’s not enough just to surface patterns in data.
In my honest opinion the disregard for the data generating process is exactly why most computer science attempts to "infer" causality fall flat on their faces. We can use causal structure to build our models, but rarely can we learn causal structure from arbitrary data.
What are we actually trying to reason about when we estimate an average treatment effect? Fixating on the estimate can mislead us whereas visualizing heterogeneity can be clarifying. Cool illustration of this idea!
Often we stop interpreting experiment results at an average treatment effect (ATE). But the same ATE can result from different patterns of variation. Causal quartets are like Anscombe's quartet, but for illustrating different patterns of variation we might see across units. 1/
Today I attended the
@DSI_UChicago
launch! It was very exciting to hear about research around essential Qs for DS as a discipline, teaching focused on both practical experience and theoretical foundations, and outreach efforts with the local community, government, and industry!
Although I am super skeptical about generative AI, I did finally decide to try ChatGPT today (in anticipation of needing it to streamline some rote tasks in the near future). I have to say, the initial feeling of awe and excitement is undeniable!
@lakens
Methodological gatekeeping is not the purview of the IRB. I worry that this sort of thing will slow down research and erode the intellectual freedom of the faculty, perhaps censoring certain paradigms. An administrative unit should not have such influence over research design.
The whole “but is this a ___ paper?” thing is almost always gatekeeping. HCI and vis are big tent fields, so try to keep an open mind folks!
Our job as reviewers is to see the best possible version of the submitted work published, to lift up our peers, and help them grow.
If you're a reviewer for VIS and you're getting started with the discussion process, I have only one request for you.
Tone down the "but is this a vis paper" thing, *please*. Nothing good ever comes out of it.
Imo narratives about AI relegating human expertise to obsolescence are alarming. This is the wrong design goal for these technologies and may result in a world where people think less and place more trust in bloated/fragile infrastructure than our own minds.
Working with Jessica and Matt was a life-changing opportunity for me. Maybe it will be for you too? Prospective PhD students, consider applying to work with these wonderful people!
I'm seeking a new PhD student at
@NorthwesternCS
! Interests include modeling visual analysis/learning from data under uncertainty, interface-mediated feedback loops, implications & decision making w/differential privacy. Join my lab w/
@mjskay
! Pls retweet!
I'm excited to share this blog post I wrote with
@JessicaHullman
about the study I'm presenting at
@ieeevis
next week. We conducted two experiments, with
@_fmnguyen
and
@mjskay
, investigating the impacts of uncertainty visualizations on trend perception.
Congrats to IDL member
@AlexKale17
,
@yifanwu
, and former IDL co-director
@JessicaHullman
(we still miss you 😢) for their
#ieeevis
Honorable Mention paper “Causal Support: Modeling Causal Inferences with Visualization”!
Seeing posts about NSF GRFP rejections, I’m reminded that I was denied funding (for work that I went on to do anyway) because one reviewer worried that I could not implement what I proposed. I suspect such judgments of who is “qualified” are poorly calibrated and launder bias.
I really dislike zero-based indexing. Maybe it’s because I switch between programming languages often. Maybe it’s because counting numbers start with 1. 🤷♂️
Am I alone in this?
#HCI
folks! No more CHI rebuttals! CHI 2022 is transitioning to revise and resubmit model!
@NElmqvist
, time to write a blog post on how to R&R for CHI 2022.
#chi2021
I’m teaching intro CS in the fall using Python. What are people’s favorite VS Code extensions and workflows for Python? Links to beginner-friendly tutorials/demos would be much appreciated!
I am profoundly disturbed by the lack of regard for human life I’m seeing:
- about the Russian war of aggression
- about the ongoing pandemic
- about the impending climate disaster
- about the rights and liberties of trans people in the US
We have to do better.
I’m so disappointed/disgusted at the repeated incidents of men harassing women in CS. The revelations about what’s been happening at UMich are part of a broader pattern of behavior. We can do so much better than this. Survivors, we hear you, we believe you, we respect you. 😤
I’m only pointing out the tip of the iceberg here. But my point is that if you’re an international scholar wondering if academics in the US support these xenophobic and exclusionary policies, know that many of us think they are a horrible and self-sabotaging injustice.
I am seeking participants for an interview study on the role of interactive visualization in participatory budgeting in the City of Chicago. If you’re an academic, policy worker, or community organizer involved in participatory budgeting, I would love to hear from you!
“It’s like the idea of an honest talk implies that you’ll lose more people, because you’ll have to tell a more complicated story than the one that you yourself were once fooled by.” -
@JessicaHullman
Pandas should not be the go to Python package for data frames.
As an R person pushed into teaching with Python (what students are familiar with), I’m thinking of teaching polars just so that my students and I don’t have to puzzle over messy pandas syntax.
@seanjtaylor
Imo overestimates should be positive and underestimates should be negative. ‘predicted - actual’ as an error function produces a more cognitively natural visual encoding: overestimate (pos) -> up; underestimate (neg) -> down (assuming standard Cartesian graphical conventions).
@MollyBAtkinson
As one of my mentors would say, “There’s always more work to do.” I triage what needs to be done. I set boundaries with people. I set expectations about what I can and cannot take on. I still often work long hours, but I mix it with leisure on the weekends to keep myself happy.
We had fun designing this interactive interface around UMAP projections. 🐙 Brushing on the scatterplot shows local distributions of features in the bars. Clicking on a bar maps that feature to opacity in the scatterplot. This helped us explore the meaning of the embedding space.
IDL’s
@mathisonian
demos tools for making sense of how people engage with interactive articles; here using dimensionality reduction to map behaviors.
#eurovis
The classic HCAI response is “augmentation, not replacement”, but I am skeptical that we will end up there in practice. Consider the potential for mental laziness, a collapse of diversity of thought/expression, or collective forgetting how to do tedious but important things.
Students can better prepare themselves for a career in a data and A.I.-driven world with new joint MBA and MS-ADS degree program
@ChicagoBooth
@UChicagoPSD
Pedro is an amazing colleague, one of many people I’m glad to work with
@UChicagoCS
. Jon’s statement here is a nice tribute to the breadth of Pedro’s research!
So wonderful to have
@plopesresearch
at
@uwdub
—a true polymath, inventor, & inspiration. Thank you for joining us & sharing your lab’s remarkable work
@UChicago
in haptics, electro-stimulation, chemicals+computers, and human-computer integration. Congrats to you & your students!
As I’m making my short list of institutions to apply to next year, how institutions respond to covid (and other crises) is a strong indicator. If I see you treat your people poorly, I won’t apply. I hope other ECRs make the same judgment. Bad policy should have downstream costs.
I am not supposed to even ask if my UNC students are vaccinated to protect their sacred right to privacy but UNC can ask me to teach 100 of them indoors for three hours a week, then go home to my unvaccinated children.
@AlbertoCairo
This does not sound like consent to me: “Ghorayshi followed her to her car, at one point standing in an open car door to prevent them from driving off, adamantly arguing for the family not to leave, not to end the conversation, and above all not to pull out of the piece.”
The relationship between regression and statistical tests was one of the last things in introductory statistics to click for me. To teach students to understand NHST without centering it in lieu of estimation, should we all be teaching from
@rlmcelreath
’s Statistical Rethinking?