Ekaterina (Kat) Cleary, Center for Integration of Science and Industry

As lead data analyst and research associate for the Center for Integration of Science and Industry, Kat Cleary often plays the role of data detective. “If you torture the data long enough,” she laughs, “it will confess.”

This relentless approach paid off for Cleary’s latest research, in which she analyzed multiple data sources to determine to what extent funding from the National Institutes of Health (NIH) contributed to the discovery of new medications. Cleary began by compiling information about NIH-funded research that isolated drug “targets”— that is, a molecule in the body (usually a protein) associated with a specific illness or disease. She identified nearly 2 million publications referencing those targets that ultimately led to drug approvals, establishing a direct link between new medications and NIH-supported research.

Her results proved groundbreaking: Every single one of the 210 drugs approved by the Food and Drug Administration between 2010 and 2016 relied on research that resulted from NIH grants, which totaled more than $100 billion. Cleary’s analysis has been featured in multiple news outlets and cited in congressional hearings about Big Pharma and drug pricing.

Read the Research

Dhaval Dave, Economics

Like Cleary, Professor Dhaval Dave’s research also pertains to pharmaceuticals. However, instead of exploring how drugs are made, he examined how existing medications are marketed to doctors.

Specifically, Dave sought to discover the extent to which “detailing” — i.e., face-to-face visits made by pharmaceutical sales reps to doctors— affects how often a particular medication is prescribed. He was given exclusive access to data sets for Famvir, an antiviral drug used to treat herpes complex infections such as cold sores, shingles and chicken pox. The data set, which covered a 24-month period, included the total number of detailing visits made and pharmaceutical samples given to 150,000 doctors, as well as the number of prescriptions those doctors wrote for both Famvir and competing drugs. Dave notes that “previous literature had explored this with aggregate data, but not at a micro level. Our data set represented the largest-ever sample of physicians to be used in a study of this type.”

Ultimately, his research demonstrated that doctors who’d had more face-to-face visits with sales reps were more likely to prescribe Famvir to their patients. Dave’s study also provided the first evidence that this increase is coming at the expense of other branded or generic alternatives, which is critical for evaluating the welfare and cost implications of such advertising.

Read the Research

Rani Hoitash, Accounting

“Accounting is the language of business,” Professor Rani Hoitash says. And like any other language, it has its own structure and subtleties of meaning. In his latest research, Hoitash determined a way to leverage XBRL, or eXtensible Business Reporting Language, to measure the accounting complexity of public companies.

“Before XBRL,” he says, “accurately measuring accounting complexity simply wasn’t possible.” Hoitash says the sheer volume and diversity of accounting information is what makes the task so complicated, because it requires firms and individual auditors to possess such a deep knowledge of accounting. What’s more, this complexity can also lead to errors in financial reporting.

Since June 2018, the Securities and Exchange Commission has required all companies to use XBRL for their financial statements and related notes. This is significant, Hoitash explains, because XBRL embeds interactive tags in financial data that can be “read” by machines instead of humans, allowing access to more of a firm’s information than was previously available.

In his study, Hoitash used XBRL concepts to create a new measure of accounting reporting complexity, which he calls ARC. Ultimately, he found that the greater a firm’s ARC, the greater the likelihood of financial misstatements, weaker controls and audit delays. Hoitash notes that auditors, investors, financial analysts, regulators and other stakeholders can use this new measure to identify financial data sets that are more likely to include inaccuracies. He frequently updates this measure and shares it on his website.

Read the Research

Kerri-Ann Sanderson, Accounting

When she’s not busy teaching or researching, Assistant Professor Kerri-Ann Sanderson grabs her camera and heads outdoors. “There’s such a science to photography,” she says, noting how a subtle shift in framing her subject “can change your whole perspective.” A similar sentiment is evident in Sanderson’s latest research, which emphasizes the need for auditors to shift their perspective from conventional financial analysis to include nontraditional Big Data sources.

“Auditors tend to look at the same financial evidence to evaluate transactions,” she explains. In her view, however, Big Data — larger, more complex data sets generated from internal and external sources and facilitated by technological innovations — can enhance the auditing process. For example, Sanderson says, “A clothing company might say it sold more winter coats than ever, generating significant profits in a season. However, the auditor could analyze weather-related Big Data from key markets and find it’s been the warmest winter on record. The evidence is not aligned.” In this way, Big Data can lead the auditor to take a different approach to evaluation.

In her study, Sanderson specifically explored how Big Data visualizations — graphic presentations of information such as word clouds, pie charts and bar graphs — affect the auditing process. Conducting experiments with 127 senior auditors from the Big 4 accounting firms, she demonstrated that, when viewed after a more traditional analysis, data visualizations can help auditors deliver a more thorough assessment.

Read the Research