💬 NLP

Research Method Usage across Academic Ages in Library and Information Science: An Empirical Study (1990-2023)

This empirical study analyzes 26,677 Library and Information Science articles from 1990 to 2023 to reveal how research method usage and diversity evolve across academic career stages, finding that mid-career scholars exhibit the highest methodological variety while theoretical approaches decline in favor of experimental and bibliometric methods.

Original authors: Chengzhi Zhang, Jiayi Hao, Yi Mao

Published 2026-06-23

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Chengzhi Zhang, Jiayi Hao, Yi Mao

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the world of Library and Information Science (LIS) as a massive, bustling kitchen where thousands of chefs (researchers) are cooking up new knowledge every year. This study is like a giant food critic who has tasted and analyzed 26,677 recipes (academic papers) published between 1990 and 2023.

The main goal of this research wasn't just to see what dishes were being made, but to figure out how the chefs' cooking styles change as they gain more experience in the kitchen. The researchers used a special metric called "academic age" to measure this. Think of "academic age" not as how old a person is, but as how many years they have been publishing their own recipes.

Here is the breakdown of what the study found, using simple analogies:

1. The Three Generations of Chefs

The researchers divided the chefs into three groups based on their "academic age":

Early-Career (The Apprentices): Chefs who have been publishing for less than 7 years.
Mid-Career (The Sous-Chefs): Chefs with 7 to 14 years of experience.
Late-Career (The Head Chefs): Chefs with more than 14 years of experience.

2. How the "Cooking Tools" Changed Over Time

In the past (the 1990s), almost everyone in the kitchen was using the same old tool: Theoretical Approaches. Imagine this as chefs just writing down recipes in a notebook without actually cooking anything. They were mostly talking about how food should taste.

But as time went on (2000s to 2023), the kitchen got noisy and active. The "notebook writing" declined, and chefs started grabbing experimental tools and bibliometric tools.

Bibliometrics is like using a super-advanced calculator to count how many people ordered a specific dish and how often it was mentioned in other cookbooks.
Experimental methods are like actually running taste tests in a lab.

The study found that Bibliometrics became the most popular tool overall, but the way different generations used them varied.

3. The "Toolbox" Diversity

One of the most interesting findings was about how many different tools each group kept in their toolbox:

The Apprentices (Early-Career): They loved Experiments. They were eager to try new, hands-on techniques.
The Head Chefs (Late-Career): They stuck to what they knew. They relied heavily on Bibliometrics and had the least diverse toolbox. They were the most specialized.
The Sous-Chefs (Mid-Career): These were the most versatile cooks. They had the most diverse toolbox, mixing experiments, calculations, and surveys. They were the most adaptable group.

4. The "Menu" (Research Topics)

The researchers also looked at what the chefs were cooking.

Late-career scholars were the heavy hitters in established, classic dishes like "Scientometrics" (studying the study of science) and "Text Mining." They had spent decades building a foundation in these areas.
Early-career scholars were also very active in "Text Mining," showing that the new generation was jumping right into the modern, tech-heavy trends.
Interestingly, while the types of dishes changed, the variety of dishes (topic diversity) was actually quite similar across all three generations. Everyone was cooking from a fairly wide menu.

5. The "Heat Map" of Preferences

If you looked at a heat map of the kitchen:

Early-career chefs were the ones most likely to use Questionnaires (asking people what they think) and Experiments when studying things like "User Behavior" or "Social Media."
Late-career chefs were the ones most likely to use Questionnaires when looking at "Library Education" or "Social Media," but they leaned heavily on Bibliometrics for everything else.

The Bottom Line

This study is like a map showing that as researchers grow older in their careers, their "cooking style" shifts.

Newcomers are eager to experiment and try new things.
Veterans rely on the powerful, data-heavy tools they mastered long ago.
Mid-career researchers are the most flexible, using a mix of everything.

The researchers conclude that understanding these shifts helps us see how the field of Library and Information Science is evolving. It shows that the discipline is moving away from just talking about ideas (theory) and toward actually measuring and testing things (empirical methods), with different generations leading the charge in different ways.

What the study did NOT do:
The paper does not claim that one group is "better" than the other, nor does it suggest that young researchers should copy the older ones or vice versa. It simply observes that these are the natural patterns of how the field has developed over the last 30 years. It also notes that because they only looked at English-language journals, they might have missed some "flavors" from other parts of the world.

Technical Summary: Research Method Usage across Academic Ages in Library and Information Science (1990–2023)

Problem Statement
While existing literature has extensively examined macro-level patterns of scholarly collaboration, performance, and content preferences, there is a need for micro-level analysis regarding how individual scholarly characteristics influence research behavior. Specifically, the relationship between a scholar's "academic age" (a proxy for career stage) and their methodological choices in Library and Information Science (LIS) remains under-explored. Prior studies suggest age-related differences in method preference (e.g., late-career scholars favoring qualitative methods), but a systematic, large-scale empirical investigation linking academic age, research topics, and method selection across the LIS field is lacking. This study addresses this gap by investigating differences in research method usage among early-, mid-, and late-career scholars to provide guidance for method selection and scholarly development.

Methodology
The study employs a large-scale empirical analysis of a corpus comprising 26,677 research articles published between 1990 and 2023 in 14 authoritative LIS journals (Q1–Q3 tiers of the Web of Science JCR). The methodology integrates natural language processing, bibliometrics, and statistical analysis through the following steps:

Data Preprocessing and Author Disambiguation:
- Non-research items (e.g., editorials, book reviews) were excluded.
- Author identity ambiguity was resolved using the OpenAlex database, assigning unique Author IDs to 53,683 distinct authors.
- Academic Age Calculation: Defined as the difference between the publication year of a specific article and the author's earliest publication year, plus one. Authors were categorized into three groups based on quartiles: Early-career (<7 years), Mid-career (7–14 years), and Late-career (>14 years). Outliers were removed using an adjusted Interquartile Range (IQR) method.
Topic Modeling:
- The Top2Vec model was applied to abstracts to identify latent research topics.
- Initial clustering yielded 29 topics, which were manually reviewed and merged based on semantic overlap into a final set of 20 distinct research topics (e.g., Scientometrics, Text Mining, Information Retrieval).
Automatic Method Classification:
- The CogFT (Cognize Full Text) model was utilized to automatically identify research methods within the full text of the articles.
- The classification framework was based on the 16-method taxonomy established by Chu and Ke (2017).
- CogFT utilizes SciBERT for segment encoding and a BERT classifier to assign method labels, achieving an F1 score > 0.85.
Statistical Analysis:
- Popularity: Measured as the average annual usage frequency of a method per author.
- Diversity: Measured using Shannon Entropy to quantify the variety of methods used within each academic age group.
- Significance Testing: A Kruskal-Wallis H test was conducted to verify statistical differences in method usage across age groups.

Key Results

Methodological Trends: Over the 33-year period, the share of theoretical approaches declined significantly, while experimental and bibliometric methods gained prominence.
Usage by Career Stage:
- Bibliometrics emerged as the most popular method overall across all groups.
- Early-career scholars exhibited the highest frequency of experimental methods, followed by bibliometrics.
- Mid-career and Late-career scholars relied most heavily on bibliometrics, followed by content analysis and experimental methods.
Diversity Findings:
- Mid-career scholars demonstrated the highest method diversity (Shannon Entropy = 2.291).
- Late-career scholars showed the lowest diversity (Shannon Entropy = 2.218).
- The Kruskal-Wallis test confirmed significant differences in method usage distributions across the three age groups ( $H = 241.8, p < 0.001$ ).
Topic-Method Interplay:
- Late-career scholars dominate established fields like "Scientometrics" and "Text Mining," heavily utilizing bibliometrics.
- Early-career scholars show strong engagement with emerging, application-oriented topics like "Text Mining," where they frequently employ experimental methods.
- Questionnaires are consistently prominent across all groups in topics related to user behavior and library education.

Significance and Contributions
The paper claims the following contributions to the LIS field and scholarly research:

Theoretical Expansion: By integrating the perspective of academic age, the study expands the theoretical framework for method inquiry. It moves beyond macro-level evolutionary trends to reveal how methodological preferences interact with individual career stages, offering a new lens for examining scholarly evolution.
Empirical Evidence of Intergenerational Dynamics: The findings complement longitudinal studies (e.g., Järvelin & Vakkari, 2022) by demonstrating that the observed "empirical turn" and methodological scattering in LIS are closely tied to academic age. Senior scholars tend to rely on traditional quantitative methods (bibliometrics), while junior scholars are more inclined toward emerging empirical methods (e.g., text mining), reflecting intergenerational transmission and innovation.
Practical Guidance:
- For Researchers: The study provides a reference for early-career scholars to make informed methodological decisions, helping them understand the strengths and application contexts of different methods relative to their career stage.
- For Institutions: The documented differences in method preferences can inform policies aimed at fostering method pluralism and encouraging diverse research practices.
- For Collaboration: Understanding the distinct methodological and topical profiles of scholars at different career stages can facilitate the formation of productive, complementary research teams.

Limitations
The authors acknowledge several limitations: the data is restricted to 14 English-language journals, potentially excluding relevant non-LIS or non-English contributions; the study identifies correlations rather than causal mechanisms; the analysis relies on static data rather than dynamic career-long trajectories; and the lack of disciplinary affiliation data means some patterns may reflect the preferences of non-LIS scholars publishing in LIS journals.