A study on this topic did find an association. A summary was first  presented by Medscape.com (a science-news outlet) with the headline, "Using smartphones on the toilet increases hemorrhoid risk".  Importantly, the original research has not yet been peer reviewed or published in a journal.  It was presented at a scientific conference called "Digestive Diseases Week 2025". 

 

One news outlet summarized the study and amped up the effect size with this language: "Using your phone on the toilet may dramatically increase risk of hemorrhoids."

 

And a substack post by a physician stuck on some causal language: "smartphone use while on toilet triggers hemorrhoids"

 

Here are some details from the Medscape summary:

• Researchers conducted a cross-sectional survey in adult patients undergoing screening colonoscopy.
• Survey questions were designed to assess smartphone habits while using the toilet; responses to Rome IV questionnaires for functional gastrointestinal disorders in adults; and other behaviors such as straining, fiber intake, and physical activity levels.
• The presence of hemorrhoids was assessed through direct visualization as documented in endoscopic reports.

a) Based on the description above, name at least four variables included in the study. Speculate on what the levels of each variable might be. 

b) What does it mean when they say it was a "cross-sectional survey"?

 

 

Next, here are some results as described by Medscape.com. The first results are frequency claims: 

d) If you've worked with Chapter 9, you'll recognize this as a multiple regression analysis. It's a little different from the examples in the book because it is predicting a binary outcome (having hemorrhoids or not). But you should still be able to identify the DV (or criterion variable) and the IVs (the predictor variables. What are they?

 

e) Does this association support the causal claim by author who wrote, ""smartphone use while on toilet triggers hemorrhoids"?   The study's result establishes covariance (why?), but does the study's method establish temporal precedence? Does it rule out all possible internal validity threats? 

 

Finally, Medscape.com reported that for the key association (the relationship between phone use and hemorrhoids), the p value was .044. A p value that is so close to the usual cutoff of 0.05 can be a sign that the results might have been (unintentionally) p-hacked. Here's the p-hacking definition from the textbook :

 

"Researchers can analyze their study’s results in a wide variety of ways. They might remove different outliers from the data, compute scores several different ways, or run a few different types of statistics. This exploratory practice has been dubbed p-hacking (Simmons et al., 2011), in part because the goal is to find a p value of just under .05, the traditional value for significance testing (corresponding to a 95% CI that does not include zero)"(Ch 14, p. 442)

 

Because this study's p-value is .044, it raises suspicions of p-hacking. Here's why. When researchers are analyzing data, they may try various statistical steps–maybe they add new covariates one at a time to their regression analysis, or omit certain outliers here and there, checking the p-value after each step. They might stop their explorations when the p-value for their key association finally falls below .05. Because the study's p-value is so close to .05, it's possible that's what happened here–though we can't be sure, just based on this journalist's summary. 

 

f) p-hacking isn't usually intentional. What steps can researchers take to avoid it? 

  

 

• presence of hemorrhoids (assessed during colonoscopy–levels were absent or present)
• smartphone habits while using the toilet (levels might have been never, sometimes, always).
• level of gastrointestinal disorders (levels might have been none, one, more than one).
• degree of straining (levels might have been never, sometimes, always).
• fiber intake (levels might have been number of high-fiber foods per day, from 0 to 10).
• physical activity (levels might have been 0 to more than 100 minutes per week)

    All variables were measured.

b) This means that they measured all variables at the same time. They didn't, for example, measure smartphone toilet habits first, and then diagnose hemorrhoids weeks or months later. 

 

c) We don't know if we can generalize because we don't know how they got their sample. I would assume that since the sample came from adults who were undergoing colonoscopy, they were not randomly selected. (Note: you should not mention sample size here, because sample size doesn't help you assess external validity.)