Research Methods in LIS: Concept, Types and Ethics; Historical, Descriptive, Experimental and Delphi Methods; Research Design, Hypothesis and Sampling; Data Collection; Data Analysis; Statistical Packages; and Report Writing, Style Manuals, Citation and Anti-Plagiarism Tools
2. Research methods: historical, descriptive, experimental and Delphi
| Method | Features | LIS use |
|---|---|---|
| Historical | Reconstructs the past from primary sources (records, letters, reports) and secondary sources; external criticism tests a source's authenticity, internal criticism its credibility and accuracy | History of a library movement or of a national library |
| Descriptive | Describes things as they are, without manipulating variables: surveys, case studies, content analysis | User surveys, studies of library use, case study of one library's automation; the most common method in LIS |
| Experimental | The researcher manipulates an independent variable and measures its effect on a dependent variable, controlling other factors, ideally with random assignment to experimental and control groups and pre-tests and post-tests | Does an information literacy course improve students' search skills compared with a control group? |
| Delphi | Developed at the RAND Corporation in the 1950s: a panel of experts answers questionnaires anonymously over several rounds, receiving controlled feedback of the group's responses between rounds, until opinions converge | Forecasting the future of academic libraries or identifying competencies for librarians |
3. Research design: problem, literature review, hypothesis and sampling
A research design is the plan for answering the research question. It begins with selecting a problem that is significant, researchable, feasible in time and cost, new and suited to the researcher's competence. The review of literature establishes what is known, identifies gaps, avoids duplication and supplies the theoretical framework and methods. The problem is then formulated as a clear statement with objectives, research questions, scope and operational definitions. A hypothesis is a tentative, testable statement about a relationship between variables: the null hypothesis (H₀) states no difference or relationship and the alternative hypothesis (H₁) states that there is one; hypotheses may be directional (one-tailed) or non-directional (two-tailed). Testing sets a significance level α (commonly 0.05); rejecting a true null hypothesis is a Type I error (probability α), and failing to reject a false null hypothesis is a Type II error (probability β).
| Sampling technique | How units are chosen |
|---|---|
| Simple random (probability) | Every unit has an equal chance, by lottery or random numbers |
| Systematic (probability) | Every kth unit from a list after a random start, where k = population ÷ sample size |
| Stratified (probability) | The population is divided into strata (such as undergraduates, postgraduates and faculty) and units are drawn randomly from each |
| Cluster and multi-stage (probability) | Whole groups (such as colleges) are sampled, then units within them, in one or more stages |
| Convenience, purposive, quota, snowball (non-probability) | Whoever is available; units chosen by judgement; fixed numbers from each category without random selection; respondents recruiting others, useful for hidden populations |
4. Methods of data collection; data analysis and interpretation
| Tool | Key points |
|---|---|
| Questionnaire | Written questions answered by respondents, printed or online; closed (fixed choices) or open questions; clear wording, logical order, and a pilot test before use; cheap for large samples but response rates can be low |
| Interview | Structured, semi-structured or unstructured; individual or focus group; rich, flexible data, but costly and open to interviewer bias |
| Observation | Participant or non-participant, structured or unstructured; records actual behaviour, such as how users search the OPAC |
| Library records | Circulation data, transaction and search logs, accession registers, gate counts, e-resource usage reports (COUNTER): unobtrusive and objective |
| Scales | Likert (1932; summated agreement ratings, commonly five points), Thurstone (1928; equal-appearing intervals), Guttman (cumulative), semantic differential (Osgood, 1957; bipolar adjectives) |
| Checklist | A list of items to be ticked as present or absent, used for evaluating services, websites or facilities |
Data analysis organises, summarises and interprets data. Data are first edited, coded and tabulated, then presented in tables and graphs: bar charts and pie charts for categories, histograms, frequency polygons and ogives for distributions, and scatter diagrams for relationships. Descriptive statistics summarise: measures of central tendency (mean, median, mode) and dispersion (range, variance, standard deviation, coefficient of variation). Inferential statistics generalise from sample to population: t-tests and ANOVA compare means, chi-square tests association between categorical variables, Pearson's r and Spearman's rho measure correlation, and regression predicts one variable from others. Worked example: five users borrowed 2, 4, 4, 6 and 9 books in a month. The mean is 25 ÷ 5 = 5, the median (the middle value when ordered) is 4, and the mode (the most frequent value) is 4; the mean exceeds the median because one heavy borrower pulls it up.
5. Statistical packages; report writing, style manuals, citation and anti-plagiarism tools
| Package | What it offers |
|---|---|
| Spreadsheets (Microsoft Excel, LibreOffice Calc) | Data entry, formulas, descriptive statistics, pivot tables and charts; the starting point for most library statistics |
| SPSS (Statistical Package for the Social Sciences) | First released in 1968, now IBM SPSS Statistics; menu-driven descriptive and inferential statistics, widely used in LIS theses |
| Bibexcel | A free tool by Olle Persson of Umeå University, Sweden, for bibliometric analysis: preparing records from citation databases, counting frequencies and co-occurrences, and exporting data for network maps |
| R | A free, open source language and environment for statistics and graphics, created by Ross Ihaka and Robert Gentleman at the University of Auckland (1993); thousands of packages, including bibliometrix for science mapping |
A research report has preliminaries (title page, declaration and certificate, acknowledgements, contents, lists of tables and figures, abstract), a main text (introduction with the problem and objectives, review of literature, methodology, data analysis and interpretation, findings and discussion, conclusions and suggestions) and end matter (references or bibliography, appendices with instruments, and sometimes an index). Its style is clear, precise and objective, with consistent tense and terminology, and it follows the guidelines of the university and, for doctoral work, the UGC's regulations. Style manuals standardise citation and presentation: the APA Publication Manual (7th edition, 2019; author-date, common in social sciences and LIS), the MLA Handbook (9th edition, 2021; author-page, humanities), the Chicago Manual of Style (18th edition, 2024; notes-bibliography or author-date), Harvard (author-date), IEEE (numbered, engineering) and Vancouver (numbered, biomedicine). Online citation tools generate formatted references; reference management tools store, organise and insert them: Zotero (free and open source, 2006), Mendeley (2008, owned by Elsevier since 2013), EndNote (Clarivate) and JabRef (for BibTeX). Anti-plagiarism tools such as Turnitin, iThenticate, Ouriginal and DrillBit compare text against databases; INFLIBNET's Shodh Shuddhi service provides such a tool to Indian universities. A report is evaluated for the clarity and significance of the problem, the adequacy of the literature review, the soundness of design, sampling and instruments (validity and reliability), the appropriateness of analysis, whether conclusions follow from the data, ethics, and presentation.
Key takeaways
- Research explores, describes, explains and predicts; ethics forbid fabrication, falsification and plagiarism and require consent and confidentiality; research is basic or applied, interdisciplinary (integrated) or multidisciplinary (side by side).
- Historical research uses external and internal criticism; descriptive research (surveys, case studies) is the most common in LIS; experiments manipulate an independent variable with control groups; Delphi (RAND, 1950s) uses anonymous experts, iterative rounds and controlled feedback.
- Design runs from problem selection and literature review to hypotheses (H₀ and H₁; Type I error rejects a true null, Type II fails to reject a false one) and sampling (random, systematic with k = N ÷ n, stratified, cluster; convenience, purposive, quota, snowball).
- Data come from questionnaires, interviews, observation, library records, scales (Likert 1932, Thurstone 1928, Guttman, semantic differential) and checklists; analysis uses tables, graphs and descriptive and inferential statistics (for 2, 4, 4, 6, 9: mean 5, median 4, mode 4).
- Packages: spreadsheets, SPSS (1968), Bibexcel (Persson, Umeå) and R (1993); reports follow a standard structure and style manuals (APA 7th 2019, MLA 9th 2021, Chicago 18th 2024, IEEE, Vancouver); reference managers Zotero, Mendeley, EndNote; anti-plagiarism tools including Shodh Shuddhi.
Practice questions (10)
Attempt each one before opening the answer. Every explanation names the tempting wrong option as well as the right one, because that is where marks are lost.
A researcher wants a systematic sample of 100 from a list of 2,000 library members. What is the sampling interval k? Type the number.
Numerical answer — type the value.
Show answer
Answer: 20
k = population ÷ sample size = 2,000 ÷ 100 = 20. After a random start between 1 and 20, every 20th member on the list is selected.Five users borrowed 2, 4, 4, 6 and 9 books in a month. What is the mean number of books borrowed? Type the number.
Numerical answer — type the value.
Show answer
Answer: 5
Mean = (2 + 4 + 4 + 6 + 9) ÷ 5 = 25 ÷ 5 = 5. The median and mode are both 4; the mean is higher because the one user who borrowed 9 books pulls it up.The Delphi method was developed at:
Show answer
Answer: D — The RAND Corporation
Delphi was developed at the RAND Corporation in the 1950s for forecasting. DRTC is Ranganathan's centre in Bangalore, the Chicago Graduate Library School pioneered research-based librarianship, and Umeå is where Bibexcel was written.Rejecting a null hypothesis that is actually true is a:
Show answer
Answer: C — Type I error
A Type I error rejects a true null hypothesis; its probability is the significance level α. A Type II error fails to reject a false null. Sampling frame and non-response errors concern who is included in a sample, not hypothesis decisions.Which of the following are probability sampling techniques? Select all that apply.
Show answer
Answer: A — Systematic sampling; C — Stratified random sampling
Stratified and systematic sampling give every unit a known chance of selection. Snowball and quota sampling select units without randomisation, so they are non-probability techniques.Assertion (A): In historical research, internal criticism examines the credibility of a source's content. Reason (R): External criticism establishes whether a source is authentic, that is, whether it is what it claims to be. Choose the correct option.
Show answer
Answer: B — Both A and R are true, but R is not the correct explanation of A
Both are true, but R does not explain A. Internal and external criticism are two distinct stages: external criticism tests authenticity (genuineness, date, authorship), internal criticism tests whether the content is accurate and credible. One statement describes each; neither explains the other.Bibexcel, a tool used in LIS research, was developed by:
Show answer
Answer: A — Olle Persson
Olle Persson of Umeå University wrote Bibexcel for bibliometric analysis. Ihaka and Gentleman created R, Garfield founded citation indexing and the impact factor, and Priem co-authored the altmetrics manifesto.Which of the following are reference management tools? Select all that apply.
Show answer
Answer: A — EndNote; B — Mendeley; C — Zotero
Zotero, Mendeley and EndNote store, organise and format citations. iThenticate is a similarity-checking (anti-plagiarism) service, not a reference manager.A scale that asks respondents to rate their agreement with statements from "strongly disagree" to "strongly agree" and adds the ratings is a:
Show answer
Answer: B — Likert scale
Likert's summated rating scale (1932) adds agreement ratings across items. Thurstone's scale uses statements with pre-assigned values at equal-appearing intervals, Guttman's is cumulative, and the semantic differential uses bipolar adjective pairs.The 7th edition of the APA Publication Manual, widely used in LIS, was published in:
Show answer
Answer: A — 2019
APA's 7th edition appeared in 2019, replacing the 6th of 2009. The MLA Handbook's 9th edition dates from 2021 and the Chicago Manual of Style's 18th from 2024.