Unit 6: Presentation of Data and Report Writing

I. Orientation

Research produces information that must be organized, displayed, interpreted, and communicated before it can support a conclusion. Data presentation converts observations into a form that reveals patterns, while report writing provides a systematic record of the research problem, method, findings, and implications. The governing principle is that presentation must remain accurate, relevant, clear, and traceable to the collected evidence.

• Purpose: Transform raw observations into understandable evidence without changing their meaning.
• Objectivity: Tables, graphs, and written conclusions must represent the data rather than the researcher’s preference.
• Comparability: Categories and measurements should use consistent definitions, units, time periods, and population boundaries.
• Clarity: A reader should identify the subject, source, unit, and main message of a presentation quickly.
• Reproducibility: Another researcher should be able to understand how data were classified, calculated, and reported.
• Ethical responsibility: Confidential information must be protected, and results must not be selectively presented or misleadingly visualized.

II. Data Classification, Tabulation and Graphical Presentation

This section concerns the movement from unorganized data to an analytical display. Raw data such as individual ages, questionnaire responses, or laboratory readings become useful when they are classified into meaningful groups, tabulated into rows and columns, and represented graphically.

A. Classification, Tabulation and Graphical Presentation of Data

Classification organizes similar observations; tabulation summarizes them numerically; and graphical presentation displays relationships visually. These stages are connected but serve different purposes.

• Classification: Group observations according to a shared characteristic.
  • Qualitative classification: Uses attributes such as gender, occupation, educational level, or type of service.
  • Quantitative classification: Uses numerical values such as age, income, test score, or production volume.
  • Chronological classification: Arranges observations by time, such as annual enrollment from 2020 to 2024.
  • Geographical classification: Groups observations by place, such as districts, states, or countries.
• Class boundaries: A continuous variable should have clearly defined, non-overlapping intervals. For example, age groups of 10–19, 20–29, and 30–39 avoid ambiguity when the convention is specified.
• Frequency: The frequency is the number of observations in a category or class. If 18 of 60 respondents select “satisfied,” the frequency is 18 and the percentage is 30%.
• Good categories: Categories should be exhaustive, mutually exclusive, relevant to the research question, and sufficiently precise for analysis.
• Tabulation: A table arranges information in a logical structure.
  • Title: States what, where, and when the table represents.
  • Headnote: Explains units, such as “Figures in percentages.”
  • Captions: Identify column headings.
  • Stubs: Identify row headings.
  • Body: Contains the numerical or textual data.
  • Source and notes: Identify the origin and clarify unusual symbols or exclusions.
• Types of tables: A simple table describes one characteristic; a two-way table cross-classifies two variables; a multiple table examines several related characteristics.
• Percentage calculation: Percentages make groups comparable when their totals differ.

Percentage = (Category frequency / Total frequency) x 100

• Worked example: If 45 out of 150 participants use an online database, the percentage is (45 / 150) x 100 = 30%. The table should state whether respondents could select more than one database.
• Bar graph: Uses separated bars for discrete categories, such as departments or preferred research methods. Equal-width bars and a labelled vertical scale are essential.
• Histogram: Uses adjoining bars for continuous class intervals, such as income groups. The absence of gaps distinguishes it from an ordinary bar graph.
• Pie chart: Shows parts of a whole. A category representing 25% occupies one-quarter of the circle, or 90 degrees.

Sector angle = (Category frequency / Total frequency) x 360°

• Line graph: Shows change over an ordered scale, especially time. Monthly expenditure from January to June can be plotted with months on the horizontal axis and expenditure in currency units on the vertical axis.
• Frequency polygon and curve: Connect class midpoints to display the shape of a distribution and compare two distributions.
• Scatter diagram: Plots paired values, such as hours of study and examination scores, to inspect direction and strength of association. It does not by itself prove causation.
• Graphical accuracy: The baseline, scale intervals, labels, legend, and measurement units must be visible. A truncated vertical axis can exaggerate a small difference.
• Selection principle: Use tables when exact values matter, graphs when patterns matter, and both when readers need pattern plus numerical verification.

B. Applications and Limitations

The value of presentation depends on whether the selected form supports the research objective and preserves the limits of the evidence.

• Application in comparison: A two-way table can compare satisfaction by age group, while clustered bars can make the same differences easier to see.
• Application in trend analysis: A line graph can show that response rates rose from 52% to 68% over four survey rounds.
• Application in distribution analysis: A histogram may reveal concentration, skewness, or unusually wide variation that an average alone hides.
• Avoiding distortion: Areas, three-dimensional effects, excessive colours, and unequal intervals can create visual impressions unsupported by the data.
• Missing information: “No response,” “not applicable,” and zero are different conditions and should not be combined without explanation.
• Interpretive limit: A graph describes the displayed data; generalization to a larger population requires an appropriate sample and statistical justification.

III. Report Writing

A research report is a formal communication of an investigation. Its purpose is to enable readers to understand what was studied, why it was studied, how evidence was obtained, what was found, and what the findings mean.

A. Report Format and Sections

A report follows a logical sequence from identification of the study to evidence, interpretation, and supporting documentation. The exact format may vary by institution, but the function of each major section remains stable.

• Title page: Gives the title, researcher’s name, institution, course or department, and submission date. A title such as “Social Media Use and Academic Performance among First-Year Students” identifies the variables and population more effectively than “A Study of Students.”
• Abstract or executive summary: Briefly states the problem, method, principal findings, and conclusion. It is written after the report is complete, although it appears near the beginning.
• Table of contents: Lists headings and page numbers so readers can locate sections efficiently. Tables, figures, and appendices may have separate lists.
• Introduction: Establishes the background and context of the study.
  • Problem statement: Defines the specific issue requiring investigation.
  • Objectives: State what the research intends to achieve, using action verbs such as determine, compare, or assess.
  • Research questions or hypotheses: Convert the problem into answerable propositions. A hypothesis might predict that training hours are positively related to performance.
  • Scope and significance: Identify boundaries and explain the study’s contribution.
• Review of literature: Synthesizes relevant theories and previous studies. It should identify agreement, disagreement, methods used, and the gap addressed by the present research.
• Methodology: Explains how the evidence was produced.
  • Research design: Identifies the overall approach, such as survey, experiment, case study, or mixed methods.
  • Population and sample: Defines the target group and selection process. A sample of 120 employees from a population of 2,000 should include the sampling method and inclusion criteria.
  • Instruments: Describes questionnaires, interview schedules, observation forms, or measurement devices.
  • Procedure: Presents data-collection steps in their actual sequence.
  • Analysis: States whether frequencies, means, tests of association, thematic coding, or other procedures were used.
  • Ethics: Records consent, confidentiality, voluntary participation, and approval where applicable.
• Results or findings: Presents analyzed evidence without extensive interpretation. Tables and figures should be numbered consecutively and introduced in the text.
• Discussion: Explains the meaning of findings, relates them to objectives and earlier research, and considers plausible reasons for agreements or differences.
• Conclusion: Answers the research questions directly and draws only conclusions supported by the findings.
• Recommendations: Converts justified conclusions into practical or research actions. Each recommendation should identify an appropriate recipient or context.
• References: Lists works cited in the required academic style. Every in-text citation should correspond to an entry.
• Appendices: Include supporting material such as the questionnaire, consent form, detailed calculations, coding framework, or supplementary tables.
• Writing conventions: Use precise headings, consistent tense, numbered pages, defined abbreviations, and a uniform citation and table style. Avoid unsupported claims such as “everyone agreed” when the sample included only 40 respondents.

B. Applications and Limitations

Effective report writing connects evidence to decisions while making the boundaries of the study visible.

• Application for decision-making: A clearly reported finding, such as “training reduced average processing time from 12 to 9 minutes,” can guide management action.
• Application for replication: A detailed method allows another researcher to repeat the study or test it in a different population.
• Limitation statement: A report should identify constraints such as small sample size, self-report bias, non-response, restricted geography, or short observation period.
• Validity of conclusions: A statistically significant association does not automatically establish causation; design, alternative explanations, and measurement quality must be considered.
• Presentation discipline: Findings should be separated from discussion so readers can distinguish observed evidence from interpretation.
• Ethical reporting: Negative, unexpected, or non-significant findings should be retained when they affect the research question. Fabrication, falsification, plagiarism, and selective omission damage the credibility of the report.

IV. AI Application in Research

Artificial intelligence refers to computational systems that perform tasks such as classification, prediction, language processing, pattern detection, or content generation. In research, AI is an assistive tool whose outputs require human verification, methodological justification, and ethical control.

A. AI Application in Research

AI can support several stages of research, but it does not replace the researcher’s responsibility for the question, evidence, interpretation, or final report.

• Literature discovery: Search and recommendation systems can identify related articles, keywords, and citation links. Researchers must verify that each source actually exists and supports the stated claim.
• Data classification: Machine-learning models can assign survey comments to categories such as “cost,” “quality,” or “access.” A manually coded sample should be used to check accuracy.
• Data cleaning: AI-assisted tools can flag duplicate records, inconsistent spellings, missing values, or unusual observations. A flagged value is not automatically an error; a legitimate extreme case may require retention.
• Pattern detection: Algorithms can identify clusters, trends, or associations in large datasets. For example, a model may detect that complaints increase during particular service periods.
• Transcription and translation: Speech-to-text systems can convert interviews into searchable text, but accents, technical terms, speaker identities, and confidential content require review.
• Analysis assistance: AI can suggest statistical procedures or code, but researchers must confirm assumptions such as independence, normality, sample adequacy, and correct variable definitions.
• Report drafting: Generative AI may help organize headings, improve grammar, or produce an initial summary from verified findings. It must not invent participants, results, citations, or quotations.
• Reproducibility: Record the tool name, version, date, purpose, prompts where relevant, data-processing steps, and human checks. This creates an audit trail for AI-assisted work.
• Bias and fairness: Training data may contain social or measurement bias. A classifier that performs well overall may misclassify a minority group, so performance should be examined across relevant subgroups.
• Privacy and security: Personal, confidential, proprietary, or identifiable research data should not be uploaded to an AI service unless approved safeguards and data-processing arrangements are in place.
• Human accountability: The researcher remains responsible for informed consent, data protection, analytical decisions, accurate reporting, and the final interpretation.

B. Applications and Limitations

AI is most defensible when it increases efficiency without obscuring how conclusions were produced.

• Appropriate use: Use AI for transcription, preliminary coding, formatting, anomaly flagging, and language editing when outputs are checked against original records.
• Inappropriate substitution: Do not treat generated text as evidence or allow an algorithm to make consequential research judgments without documented criteria and review.
• Accuracy limitation: AI systems can produce plausible but false statements, sometimes called hallucinations; every factual claim and citation requires source verification.
• Transparency requirement: State material AI use in the methodology or acknowledgements according to institutional policy, especially when it affects analysis or interpretation.
• Quality control: Compare AI classifications with human labels using measures such as accuracy or agreement, then revise categories when systematic errors appear.
• Final principle: AI may accelerate research operations, but valid knowledge still depends on sound design, trustworthy data, transparent analysis, and responsible human judgment.