abstracts analyzed
22,179
name a quantitative method
30%
rise in PLS-SEM since 2021
2×

Which statistical methods do business dissertations actually use? We collected every English-language dissertation and thesis with an abstract published since 2021 in five areas of business (22,179 abstracts) and counted which methods they name. As in our psychology and education series, “dissertations” is short for dissertations and theses.

This post is the overview. Each area has its own breakdown: strategy and management, accounting and finance, marketing, organizational behavior and HRM, and information systems and technology management.

Fewer than a third of business abstracts name a method

30% of business abstracts (6,650) name a quantitative method. Another 21% describe qualitative work, 14% a survey, and 12% a case study. Experiments are rare (1%).

Bar charts by business area. Share naming a quantitative method: strategy 24%, accounting and finance 37%, marketing 35%, OB and HRM 30%, information systems 21%. Qualitative: 26%, 9%, 21%, 33%, 28%. Survey or questionnaire: 11%, 8%, 19%, 26%, 14%. Experiment: 1%, 1%, 2%, 2%, 1%.
Design language by area, as a share of all abstracts. The dashed line marks business overall.

Accounting and finance is the most quantitative area (37%) and the least qualitative (9%). Organizational behavior and HRM leans most on surveys (26%), and information systems and strategy lean on case studies (18% and 17%).

Two toolkits

Among the 6,650 method-naming abstracts, linear regression leads (18%), followed by correlation (15%), structural equation models (12%), reliability (12%), moderation (12%), and mediation (11%). Split by area, two toolkits appear.

Heat map of 14 methods across five business areas. Marketing stands out for SEM (24%) and PLS-SEM (24%); OB and HRM for mediation (28%), moderation (18%), and SEM (17%); information systems for machine learning (22%); accounting and finance for panel data (15%), causal designs (12.5%), and time series; strategy for panel data (15%) and moderation (16%).
Each cell is the share of that area’s method-naming abstracts that name the method. Outlined cells are clearly more common than in the other four areas.
  • The survey-model toolkit (marketing, organizational behavior, information systems): questionnaires, reliability, and structural equation models, often PLS-SEM, with mediation and moderation at the center. Organizational behavior names mediation in 28% of method-naming abstracts; marketing names SEM and PLS-SEM in 24% each.
  • The econometric toolkit (accounting and finance, strategy): archival or financial data analyzed with panel models (15%), difference-in-differences, instrumental variables, and regression discontinuity (7.5–12.5%), and time-series models.

Machine learning cuts across both, and is most common in information systems (22%).

What changed since 2021

Dumbbell chart comparing 2021–22 with 2025–26 in business. Reliability rose from 2.4% to 4.5% of abstracts, SEM from 2.6% to 4.2%, mediation from 2.6% to 3.8%, machine learning from 2.6% to 3.9%, panel data from 2.1% to 3.6%, PLS-SEM from 1.6% to 3.4%, and causal designs from 1.6% to 2.5%.
Share of all abstracts naming each method, 2021–22 vs. 2025–26. Coral: rises that survive correction.

Business became more quantitative: 27% of abstracts named a method in 2021–22 and 33% in 2025–26. Both toolkits grew. On the survey side, PLS-SEM more than doubled (1.6% to 3.4% of all abstracts), and SEM, mediation, moderation, and reliability all rose. On the econometric side, panel models rose from 2.1% to 3.6% and causal designs from 1.6% to 2.5%. Machine learning rose from 2.6% to 3.9%. ANOVA and data envelopment analysis became less common.

US-tagged work is similar, with one big exception

Dumbbell chart comparing US-tagged business abstracts with all others. Most methods are similar. PLS-SEM appears in 2% of US-tagged method-naming abstracts versus 10% elsewhere, and network analysis in 5% versus 1%.
Share of method-naming abstracts, US-tagged (coral) vs. other or unknown country (open).

Only 3% of these works are tagged as US institutions (country is known for 14%), so the US sample is small: 200 method-naming abstracts. Most methods look similar, with two exceptions. PLS-SEM appears in 2% of US-tagged method-naming abstracts, compared with 10% elsewhere, and network analysis in 5% compared with 1%. PLS-SEM is popular in many business programs, but many reviewers question it for confirmatory theory testing. If you plan to use it, be ready to explain why.

What abstracts leave out

Four of 22,179 abstracts mention a power analysis, and 14 mention missing data. Archival studies rarely need a power analysis, but survey studies do, and both need a plan for missing values and outliers.

That fits APA’s own guidance. Its reporting standards ask a quantitative abstract for the research design, sample size, measures, and findings with effect sizes, not the name of the statistical test or the power analysis; those belong in the Method section. Our guide to writing a dissertation abstract covers what to include, with data on what real abstracts leave out.

What this means if you're planning a dissertation

  • Know which toolkit your area expects. A marketing committee and a finance committee will ask very different questions about the same data.
  • Justify your SEM choice. Covariance-based SEM and PLS-SEM answer different questions. Pick the one that fits your goal, and report measurement quality first.
  • Say what makes your estimate causal, or don't claim it is. Panel fixed effects, difference-in-differences, and instrumental variables each rest on assumptions you should test and report.
  • Treat survey mediation carefully. A single cross-sectional survey can show a pattern consistent with mediation, not prove it.

Read the area breakdowns

How we did this

We used OpenAlex to collect every English-language work typed as a dissertation or thesis, with an abstract, published from 2021 through September 2026, in six OpenAlex subfields: strategy and management, accounting, finance, marketing, organizational behavior and human resource management, and management information systems together with management of technology and innovation. We kept the 22,179 abstracts longer than 60 words and searched each title and abstract for about 40 methods, including business and econometric methods such as panel models, difference-in-differences, PLS-SEM, and data envelopment analysis. We spot-checked matches and tightened patterns that caught false matches, such as “VaR” (value at risk) counted as vector autoregression and “computer-mediated” counted as mediation.

Keep these limitations in mind. Percentages describe how often abstracts mention a method, not how often it was used. The data are international and include master's theses; country is known for only 14% of works. OpenAlex assigns subfields by algorithm, so boundaries are loose. OpenAlex's coverage changes over time, which can affect trends. Comparisons use Fisher's exact tests with a Benjamini–Hochberg false-discovery-rate correction across all methods tested.

The analysis code, the method dictionary, and the result tables behind every number here are on GitHub, under the MIT License.