2026 Statistics PhD Survey

In 2026, we conducted an opt-in survey of 66 folks who have participated in a Statistics PhD. The majority of survey respondents have already graduated, but there are a few current students. The goal was to understand sentiment and opinions on today's relevant topics.

We also acknowledge that there are issues with a survey of this type: namely that it's a non-random sample, and only has 66 responses. Nevertheless, the trends are still interesting and provide insights into key trends.

This post summarizes the primary findings of the survey. If you'd like to explore more, you can view an interactive visualization of the results at this link. With that, let's get into the results.

Most people are satisfied with the PhD, would do it again, and would recommend it to others

Most survey respondents were satisfied with the PhD. Specifically, ~90% said they were either 'Satisfied' or 'Extremely satisfied':

We unfortunately do not have a good benchmark, for example against Math PhDs. The only other survey I could find was the Survey of Earned Doctorates. But this was more tailored to general tracking.

Another data-point that indicates general satisfaction is that ~80% of respondents said that they would do their PhD again:

And ~77% said that they would recommend a PhD to others:

Many survey respondents highlighted that their recommendation was nuanced. This is well encapsulated in the following response:

"It depends on what their goals are. I think that on average, the recommendation should be No. It's really a 'satisfy your curiosity' kind of thing — the vast majority of people will end up working in non-academic roles, and in those roles, hands on experience in industry leads to faster career growth than doing a PhD."

— Data scientist in tech

Satisfaction varies by graduation year, department, and industry

There is considerable variation in satisfaction within different groups of survey respondents. Most notably, satisfaction declines by graduation year:

There are a few hypotheses for why this is the case.

The hyfirst is recency: people who just graduated may be more negative about their experiences, compared against those with more time to reflect. It could also be a real decline. Another hypothesis is that satisfaction dipped during COVID and the rise of remote work, which means it could rebound over time. We unfortunately can't test these because we do not have multiple responses from the same person over time, which highlights the importance of repeating this survey over the years.

Satisfaction also varies by department. About ~70% of survey respondents went to CMU, and satisfaction from CMU tended to be higher than other departments:

The final major variation in satisfaction is post-PhD industry. Folks who stay in Academia have higher satisfaction, compared against those who land in other industries:

One potential explanation for this is that, for professors, a PhD is directly linked to their current job:

You need to survive and thrive in the PhD environment to even have a chance to be a faculty running your own lab.

— Current biostatistics professor

This contrasts with industry respondents, where a PhD is not really necessary once you get a foot in the door:

Unlikely to help much for industry positions; the only people who should do a PhD are the people who will ignore all advice to not do one.

— Machine learning engineer in tech

There are major differences by post-PhD industry

The differences in responses based on post-PhD industry are perhaps the most interesting. For example, the following chart identifies what was most useful about the PhD:

Respondents who work in Tech were much more likely to find networking with peers, research with peers, and applied courses useful. In contrast, academics were more likely to value networking outside their cohort, doing research with other faculty, and theoretical courses. Though notably, applied courses were deemed more useful than theoretical courses for acadmics and tech folks alike. Each group agreed that research with their advisor was the most useful part of the PhD.

There is advice to be gleaned here for future PhD students. If you plan to go to industry, then your peers are the most important people to get to know. In academia, it is more important to mingle outside your cohort, either by doing research with other professors, or simply making connections for a future job search.

AI has the most impact in Tech, and Academics are the most concerned

~70% of tech industry respondents see AI having a major impact on their day-to-day job is tech, which is the highest. Academia is most likely to forecast a major impact on the label market, with 71% of respondents saying it will have a major impact:

Academia's view on AIs and the labor market is related to their forecast of future hiring for statistics PhDs:

Some of this angst is well encapsulated with the following quotes, which emphasize that Academia is focused on hiring folks versed in AI:

Academia appears to be recruiting individuals who specialize in "AI". This means that computer scientists are pouring into the field. Right now, statisticians have to label themselves as "AI people", though the task remains the same: develop flexible models for complex data. After the AI bubble bursts, things will come back to normal.

— Recently on the academic job market

Everyone will want to hire people with "AI" PhDs instead, even if "AI" actually means things that statistics PhDs are good at.

— Current statistics professor

Industry folks are more open to switching jobs

Another major difference between industry and academia is their approach to the job search. The following table breaks down the job status of survey respondents:

The tech folks are most open to switching jobs: 55% are passively open to offers (though very few are actively looking). Some of this may be tied to location, as industry concentrates in big cities. It may also be the nature of the Tech industry. This stack overflow survey shows similar figures.

There is another bit of useful advice here. If you enter academia, your will probably be either locked in on your your current job, or actively seeking new jobs. Whereas in industry, especially tech, you will be more attuned to the market, and switching jobs will be less rare.

People like the industry they chose

When you look at the industries folks find appealing, it is not at all surprising that folks are attracted to the field they are in:

95% of tech people find the tech appealing. And their second and third choices are corporate research lab (e.g. something like DeepMind), and finance. All of these are staying within industry. The high appealing rating in Finance and National Labs within the 'Other' happens because many folks in that category work in Finance, or a National Lab.

In contrast, 91% of folks in academia find academia appealing, following by tech, national lab, and corporate labs. This also makes sense, Tech is perceived as the most appealing non-research industry by academics, but they are also attracted to jobs where they can continue publishing (e.g. corporate research lab, or a national lab).

Professor is the most prestigious job, Research Scientist tops industry

Job Title is always a hot button issue. We asked survey respondents how they viewed the prestige of various job titles:

Across all groups, Professor rated as the most prestigious title. One hypothesis for this result is that all the survey respondents went through a PhD, which means they are more familiar with the effort involved to succeed in that particular career.

In tech, a close second was Research Scientist; 66% of respondents labeled it a top three most prestigious title. Research Scientist is also #2 for Academia. Although across other industries, it is ranked lower than Machine Learning Engineer, and Quantitative Researcher. Entrepreneurs have a consistent, stable level of support across industries. One other surprise finding was that 'Data Scientist' is much more prestigious in Academia than it is in industry.

Tech industry job titles are idiosyncratic and constantly shifting

The tech industry in particular, likely due to how new it is, has a very distinct taste in job titles. In particular, the Applied Scientist role is rising in prestige, beating out Data Scientist. The tech industry also places a higher value on the Machine Learning Engineer.

The perception of minute differences in job title within Tech may be due to the fact they perceive job title to be more important:

What explains this? Coupled with the higher likelihood of a job switch, folks in Tech seem to believe that the job title is functionally useful, for instance in standing out to recruiters, and finding a new job:

Prestige points seem to count, at least for getting your foot in the door.

— Research scientist

While tech workers note the utility of the title, they also indicate that they 'wish it didn't matter'

Resume first impressions matter a lot when trying to stand out among thousands of applicants. I do think that the work content is the most important, but title unfortunately matters.

— Research scientist

Other folks highlighted that having a specific job title is important when pivoting roles:

Extremely important especially when pivoting; it's hard to transition roles.

— Machine learning engineer in tech

And although the prestige of various jobs varies, respondents were able to clearly distinguish jobs that had higher prestige, against jobs that are more helpful for their overall career advancement.

Here Data Scientist, being an established job ladder across various fields at this point, regains some value. But if you drill down into the tech industry, you can see that DS doesn't appear any more helpful than the newly established roles, such as Applied Scientist, or Research Engineer. It could be a sign of either a shift, or a sense that these roles are fairly interchangeable:

Job titles have sort of coalesced around data scientist, quantitative researcher, applied scientist, and machine learning engineer. I think people have become aware of how interchangeable these titles can be.

— Data scientist in finance

Finally, it is important to note that job title prestige depends to a large extent on the company, which is well encapsulated in the following quote:

Yes, but only if tied to the affiliation. Being a professor at University X is different than University Y, and same for a Data Science Job.

— Data scientist in tech

Data Science prestige still exists, but it is waning

It was just 2012 when Data Scientist was deemed the sexiest job of the 21st century. But a major trend identified in this survey is that, at least for Statistics PhDs, it is no longer the highest prestige post-PhD job. This trend feels particularly important for the Statistics discipline, which had a major hand in the rise of data science, albeit somewhat under duress.

One way to look into this trend is by plotting a moving average of the proportion of folks who find Data Scientist one of the top three prestigious jobs, by graduation year:

Recent graduates, who it's fair to assume are most attuned to the job market, are even less likely to see Data Scientist as a prestigious job.

The shift from Data Scientist to Research Scientist may partly be explained by the AI Wave, which has even more bravado than Big Data. At AI companies, Research Scientist is the main job title if you're working at the frontier of LLMs. So it makes sense that recent graduates could be looking to break into the new industry.

The other explanation highlighted by multiple respondents is that the Data Scientist job has many new entrants, with very little barrier to entry:

The "sexiness" of a job title has always been extremely important (overly so) for one's future career moves. "Data Scientist" was once that sexy job title — infamously treading all over "Statistician". And now "data scientists" are a dime a dozen, as everyone and their mother have played with some spreadsheets, taken a CodeAcademy course, etc. and market themselves to the world as a "data scientist".

— Independent consultant

This is related to the trend in title inflation, where more groups within an organization brand themselves as Data Scientists:

It does seem like there is an implicit (all but explicit) pecking order, such that you might get taken more seriously applying to a next job if you're a Research Scientist (which is the future!) than if you're a Data Scientist (which thanks to title inflation is a title that many Analysts have).

— Data scientist in tech

Conclusion

This pilot study has been a fascinating window into the sentiment and opinions of Statistics PhDs. We summarized key takeaways in this blog post, but you can also view the results in an interactive app at the following link: 2026 Statistics PhD Survey.