Data Analysis & Results Presentation

Data Analysis and Results Presentation: Complete Guide for Quantitative & Qualitative Research (2026)

By

Last Updated: March 26, 2026

(Module 4) Data Analysis and Results Presentation:

From Concept to Submission: A Complete Guide to Research Paper and Thesis Writing

Academic Writing Mastery: The Complete 2026 Guide To Research Papers, Thesis & Dissertation Writing

Data Analysis and Results Presentation

Why Data Analysis Determines Research Credibility

Here’s what separates published research from rejected manuscripts: appropriate, rigorous data analysis. You can collect perfect data, but analyze it wrong and your conclusions are worthless.

Think about it: Your entire dissertation rests on your analysis. Thesis examiners focus heavily on this section. Journal reviewers with statistical expertise will scrutinize every test, every p-value, every interpretation.

This comprehensive guide teaches you how to analyze and present research findings correctly:

  • Quantitative analysis (descriptive and inferential statistics)
  • Qualitative analysis (coding, themes, interpretation)
  • Mixed methods integration
  • Results presentation (tables, figures, narrative)
  • Case law analysis (for law students)
  • Common analysis mistakes to avoid

Whether you’re running regressions, coding interviews, or analyzing judicial decisions, mastering analysis is essential for research success.

Understanding Your Data First

Before analysis, understand what kind of data you have.

Types of Variables

Categorical (Qualitative):

  • Nominal: Categories with no order (gender, discipline, yes/no)
  • Ordinal: Ordered categories (education level, Likert scales)

Numerical (Quantitative):

  • Interval: Equal intervals, no true zero (temperature in Celsius, IQ scores)
  • Ratio: Equal intervals with true zero (age, income, test scores)

Why this matters: Variable type determines which statistical tests you can use.

Quantitative Data Analysis

Step 1: Data Cleaning and Preparation

Before analysis:

Check for errors:

  • Impossible values (age = 250, Likert scale response = 7 when scale is 1-5)
  • Inconsistent entries (sometimes “male,” sometimes “M”)
  • Missing data patterns

Handle missing data:

  • Listwise deletion: Remove cases with any missing data (if <5% missing)
  • Mean imputation: Replace missing values with mean (if missing at random)
  • Multiple imputation: Statistical estimation of missing values (advanced)

Code variables appropriately:

  • Assign numbers to categories (Male=1, Female=2)
  • Create new variables if needed (combine items into scales)

Create codebook: Document what each variable represents and how it’s coded.

Step 2: Descriptive Statistics

Purpose: Summarize and describe your data.

For numerical variables:

  • Mean: Average (use when data is normally distributed)
  • Median: Middle value (use when data is skewed)
  • Standard Deviation (SD): Spread around mean
  • Range: Minimum to maximum

Example reporting:

“Participants’ mean age was 18.3 years (SD=0.8, range=17-21).”

For categorical variables:

  • Frequencies: Count in each category
  • Percentages: Proportion in each category

Example reporting:

“Of 450 participants, 267 (59.3%) were female and 183 (40.7%) were male.”

Step 3: Inferential Statistics

Purpose: Test hypotheses and draw conclusions beyond your sample.

Common Statistical Tests

Comparing Two Groups:

Independent t-test:

  • Compare means between two independent groups
  • Example: Do intervention students have higher retention than control students?

Reporting:

“Retention intentions were significantly higher in the intervention group (M=4.2, SD=0.7) compared to controls (M=3.6, SD=0.9), t(448)=7.3, p<.001, d=0.74.”

Key components:

  • Means and SDs for both groups
  • Test statistic (t value)
  • Degrees of freedom (df=448)
  • p-value (probability)
  • Effect size (Cohen’s d=0.74, medium-large effect)

Paired t-test:

  • Compare means for same group at two time points
  • Example: Pre-test vs. post-test scores

Comparing Three or More Groups:

One-way ANOVA:

  • Compare means across 3+ groups
  • Example: Do retention rates differ across Arts, Science, and Commerce students?

Reporting:

“Retention rates differed significantly across disciplines, F(2, 447)=12.4, p<.001, η²=.05. Post-hoc tests showed Science students (M=85%) had higher retention than Arts (M=78%) and Commerce (M=79%).”

Testing Relationships:

Correlation:

  • Measure strength and direction of relationship between two variables
  • Pearson’s r for linear relationships (-1 to +1)

Example:

“Peer support was positively correlated with retention intentions, r(448)=.54, p<.001, indicating that students with stronger peer networks had higher intentions to persist.”

Multiple Regression:

  • Predict outcome variable from multiple predictors
  • Shows which variables matter most

Example:

“Multiple regression revealed that peer support (β=.38, p<.001), academic self-efficacy (β=.26, p<.01), and financial security (β=.18, p<.05) significantly predicted retention intentions, R²=.42, F(3,446)=108.2, p<.001.”

Testing Independence:

Chi-square:

  • Test relationship between two categorical variables
  • Example: Is retention (stayed/left) related to intervention (yes/no)?

Reporting:

“Retention rates differed significantly by condition, χ²(1, N=450)=8.92, p=.003. Students in the intervention group showed 82% retention compared to 71% in the control group.”

Understanding P-Values and Significance

p-value: Probability of getting your results if the null hypothesis is true.

Convention:

  • p<.05: Statistically significant (less than 5% chance results are due to random chance)
  • p<.01: Highly significant
  • p<.001: Very highly significant

Critical: p-value alone doesn’t tell you if effect is meaningful. Always report effect sizes.

Effect Sizes

Why they matter: Statistical significance tells you IF there’s an effect. Effect size tells you HOW BIG.

With large samples, tiny differences can be “significant” but meaningless.

Cohen’s d (for t-tests):

  • 0.2 = small effect
  • 0.5 = medium effect
  • 0.8 = large effect

Eta-squared (η²) for ANOVA:

  • .01 = small
  • .06 = medium
  • .14 = large

R-squared (R²) for regression:

  • Proportion of variance explained
  • .10 = 10% of variance explained

Software Options

SPSS:

  • User-friendly, point-and-click interface
  • Common in social sciences, education, psychology
  • Expensive but many universities provide licenses

R:

  • Free, open-source
  • Powerful, flexible
  • Steeper learning curve
  • Excellent for advanced statistics

Stata:

  • Common in economics, public health
  • Good for panel data and econometrics

Excel:

  • Basic statistics only
  • NOT recommended for serious research

Qualitative Data Analysis

Step 1: Familiarize Yourself with Data

Before coding:

  • Read all transcripts or field notes thoroughly
  • Make initial notes and impressions
  • Identify potential patterns

Step 2: Coding Data

Coding: Assigning labels to segments of data.

Open coding:

  • Read through data line-by-line
  • Assign codes to meaningful segments
  • Be descriptive (codes = short labels summarizing content)

Example:

Transcript excerpt: “My mentor really helped me when I was struggling with chemistry. I would have quit without her.”

Codes: [academic support], [peer mentoring], [persistence factor]

Develop codebook:

  • Code name
  • Definition
  • When to use
  • When NOT to use
  • Example

Step 3: Identify Themes

Themes: Patterns across codes that capture important aspects of data.

Process:

  1. Group related codes together
  2. Identify overarching themes
  3. Define each theme clearly
  4. Find representative quotes

Example analysis:

Theme 1: Peer Mentors as Safety Nets

  • Definition: Mentors provide critical support during academic or personal crises
  • Codes included: academic support, emotional support, crisis intervention
  • Prevalence: 27 of 30 interviews

Representative quote:

“When I was failing math and wanted to drop out, my mentor spent hours helping me every night. Without her, I wouldn’t be here” (Participant 7, female, Science).

Step 4: Ensure Trustworthiness

Strategies:

Member checking:

  • Share findings with participants
  • Ask if your interpretation resonates
  • Revise based on feedback

Triangulation:

  • Multiple data sources (interviews + observations + documents)
  • Multiple researchers coding independently
  • Compare interpretations

Thick description:

  • Provide rich, detailed descriptions
  • Enough context for readers to assess transferability

Audit trail:

  • Document your analytical decisions
  • Keep record of code development
  • Show your reasoning process

Qualitative Software

NVivo:

  • Industry standard
  • Organize transcripts, code data, query patterns
  • Expensive but powerful

ATLAS.ti:

  • Similar to NVivo
  • Strong for theory building

MAXQDA:

  • User-friendly
  • Good for mixed methods

Free options:

  • Dedoose (cloud-based, affordable)
  • Manual coding (highlight printouts, use spreadsheets)

Important: Software helps organize, but YOU do the analysis. Software doesn’t interpret meaning—you do.

Mixed Methods Analysis

Integration Strategies

Merge:

  • Combine quantitative and qualitative databases
  • Compare findings side-by-side

Example table:

FindingQuantitative EvidenceQualitative Evidence
Peer support predicts retentionr=.54, p<.00127/30 participants cited mentor support as critical

Connect:

  • Use quantitative results to inform qualitative sampling
  • Example: Survey identifies high vs. low retainers → Interview both groups to understand why

Embed:

  • One method supports the other
  • Example: Brief interviews explain unexpected quantitative findings

Presenting Results Effectively

Narrative Results Section

Structure:

1. Restate purpose (1 sentence)

“This study examined the impact of peer mentoring on first-year retention.”

2. Present findings systematically

  • Organize by research question or hypothesis
  • Quantitative → qualitative → integration (for mixed methods)

3. Use clear topic sentences

“Hypothesis 1 predicted that students receiving peer mentoring would show higher retention than controls. This hypothesis was supported.”

4. Report statistics completely

  • Always include: test statistic, degrees of freedom, p-value, effect size
  • Example: t(448)=7.3, p<.001, d=0.74

5. Refer to tables/figures

“As shown in Table 1, retention rates were consistently higher in the intervention group across all three colleges.”

Creating Effective Tables

Good table characteristics:

  • Clear, descriptive title
  • Column headers well-labeled
  • Numbers aligned properly
  • Include sample sizes (n)
  • Note statistical significance (* p<.05, ** p<.01)

Example table:

Table 1
Retention Rates by Condition and College

CollegeIntervention n (%)Control n (%)χ²p
College A82 (83%)35 (70%)4.12.042*
College B86 (86%)39 (78%)2.34.126
College C78 (78%)33 (66%)3.45.063
Total246 (82%)107 (71%)8.92.003

Note. n = sample size. * p<.05, ** p<.01.

Creating Effective Figures

When to use:

  • Bar charts: Compare categories
  • Line graphs: Show trends over time
  • Scatter plots: Show relationships
  • Box plots: Show distributions

Figure requirements:

  • Descriptive caption
  • Labeled axes with units
  • Legend if needed
  • High resolution
  • Black and white friendly (for print)

Common Analysis Mistakes

Mistake 1: Ignoring Assumptions

Problem: Using tests without checking if data meets requirements.

Example: Using t-test when data is severely skewed (violates normality assumption)

Solution: Check assumptions first. Use non-parametric alternatives if assumptions violated.

Mistake 2: P-Hacking

Problem: Running many tests until finding p<.05, only reporting significant ones.

This is misconduct. It inflates false positives.

Solution: Pre-register hypotheses. Report all tests run, not just significant ones.

Mistake 3: Confusing Correlation with Causation

Problem: Finding correlation, claiming causation.

Wrong – “Peer support causes retention” (from correlational data)
Right – “Peer support is associated with retention” OR “Peer support predicts retention”

Only experiments can establish causation.

Mistake 4: Cherry-Picking Qualitative Quotes

Problem: Only presenting quotes that support your interpretation.

Solution: Present data representing the full range of responses. Note exceptions and contradictions.

Mistake 5: Over-Interpreting Small Samples

Problem: Claiming “students feel…” based on 5 interviews.

Solution: Qualify claims appropriately. “These 5 participants described…” or “In this small sample…”

Writing the Results Section

Results vs. Discussion

Results section:

  • Right –What you found
  • Wrong –What it means (that’s Discussion)

Example:

Results (correct):

“Retention rates were significantly higher in the intervention group (82%) compared to controls (71%), χ²(1, N=450)=8.92, p=.003.”

NOT in Results:

“This shows that peer mentoring is an effective retention strategy that colleges should implement.” (This interpretation belongs in Discussion)

Results Checklist

Before submitting, verify:

  • All research questions addressed
  • Statistical tests reported completely
  • Tables/figures referenced in text
  • No interpretation (save for Discussion)
  • Sample sizes clear
  • No contradictions with methods section

Legal Research and Writing: Complete Guide for Law Students and Legal Researchers

Conclusion

Data analysis transforms raw information into meaningful findings. Whether analyzing numbers, interview transcripts, or judicial decisions, systematic analysis is essential.

  • For quantitative researchers: Choose appropriate statistical tests, check assumptions, report completely, interpret cautiously.
  • For qualitative researchers: Code systematically, identify themes carefully, ensure trustworthiness, present richly.
  • For law students: Analyze case law to identify principles, trace evolution, synthesize doctrine; use empirical methods for socio-legal research.

For everyone: Analysis quality determines research credibility. Master these methods—your research depends on it.

Q: What is data analysis in research?

Data analysis is the systematic process of examining, cleaning, transforming, and modelling data to discover patterns, test hypotheses, and draw conclusions. For quantitative research, it involves statistical procedures — descriptive and inferential statistics. For qualitative research, it involves interpretive procedures — coding, theme development, and synthesis. The goal in both cases is to move from raw data to findings that answer the research question. Analysis must be planned before data collection, not invented after.

Q: What is the difference between data analysis and data interpretation?

Data analysis is what the data shows — patterns, frequencies, statistical relationships, themes. Data interpretation is what those patterns mean — why they occurred, what they imply for theory and practice, how they connect to existing knowledge. Analysis belongs in the Results section; interpretation belongs in the Discussion. Mixing the two is the most common structural error in research papers. Report first, interpret second — always in separate sections.

Q: How do you choose the right data analysis method?

Your analysis method is determined by your research question, research design, and data type. For quantitative data asking about relationships: correlations or regression. For group differences: t-tests or ANOVA. For categorical outcomes: chi-square or logistic regression. For qualitative data exploring experience or meaning: thematic analysis, grounded theory, or discourse analysis. The method must be specified in your methodology chapter before data collection — choosing after seeing the data introduces bias.

Q: How do you present data analysis findings in a research paper?

Present findings in the Results section without interpretation. For quantitative findings: report descriptive statistics first, then inferential test results with effect sizes; use tables for complex data and figures for trends. For qualitative findings: organise around themes, present each with analytical prose and supporting quotes. For legal research: present doctrinal findings as structured legal propositions, not summaries. Every table and figure must be referenced in text before it appears.

Q: What software is used for data analysis in research?

Common quantitative analysis software: SPSS (widely used in social science, psychology, education); R (free, flexible, increasingly preferred for advanced analysis); Stata (economics and health sciences); Excel (basic descriptive statistics only). Common qualitative analysis software: NVivo (most widely used for qualitative coding); Atlas.ti; MAXQDA. For legal research: SCC Online and Manupatra for case law analysis; no specialist software replaces legal reasoning for doctrinal analysis.

Author

Dr. Rekha Khandelwal, a legal scholar and academic writing expert, is the founder of AspirixWriters. She has extensive experience in guiding students and researchers in writing research papers, theses, and dissertations with clarity and originality. Her work focuses on ethical AI-assisted writing, structured research, and making academic writing simple and effective for learners worldwide.

Author Profile Dr. Rekha Khandelwal | Academic Writer, Legal Technical Writer, AI Expert & Author | AspirixWriters

References


Part of: Complete Research Writing Guide Series

Module 1 – The Complete Guide To Research Paper Structure: IMRAD Format, Thesis Organization & Academic Writing (2026)

Module 2 –The Academic Writing Process: Complete Guide from First Draft to Submission (2026)

Module 3 – Research Methodologies: Complete Guide To Quantitative, Qualitative, Mixed Methods & Legal Research (2026)

Next in Series:

Related Posts

Scroll to Top