Skip to the content
How to use this manual: Each experiment is a small survey project. Design the questionnaire and collect data as the steps under Formula describe, then record your results in the blank working table. The Calculation section shows an illustrative worked example (with sample summary statistics) demonstrating the primary test; replace it with your own data. The Result interprets the finding. Analyse with R, SPSS or MS Excel.

List of Practical Experiments

  1. Gender discrimination in private vs. government sector.
  2. Unemployment duration and mental health.
  3. Impact of subsidy removal on household expenditure.
  4. Online shopping behaviour among college students.
  5. Customer satisfaction with public transport.
  6. Awareness and usage of digital payment systems.
  7. Study habits and academic performance.
  8. Food consumption patterns and nutritional awareness.
  9. Impact of work-from-home on employee productivity.

Experiment 1 — Gender Discrimination: Private vs. Government Sector

1. Problem

Is there a significant difference in perceived gender discrimination between private-sector and government-sector employees?

2. Aim

To test \(H_0:\mu_{\text{private}} = \mu_{\text{govt}}\) against \(H_1:\mu_{\text{private}} \ne \mu_{\text{govt}}\) for the mean discrimination score.

3. Formula

\[ t = \frac{\bar x_1 - \bar x_2}{s_p\sqrt{\tfrac1{n_1}+\tfrac1{n_2}}}, \qquad s_p^2 = \frac{(n_1-1)s_1^2 + (n_2-1)s_2^2}{n_1+n_2-2} \]

Applying it:

  1. Questionnaire: screening (sector); a 10-item Likert discrimination scale (1–5, reverse-code positive items) giving a total score of 10–50; career progression; demographics.
  2. Sampling: stratified random — 50 private + 50 government = 100 respondents.
  3. Analysis: descriptive statistics per sector; two-sample \(t\)-test on the total score; supporting chi-square (sector × experienced discrimination) and logistic regression.

4. Calculation

Blank working table:

Sector\(n\)MeanSD
Private
Government
t / df / p

Illustrative sample statistics: Private \(n_1 = 50,\ \bar x_1 = 32,\ s_1 = 6\); Government \(n_2 = 50,\ \bar x_2 = 28,\ s_2 = 5\).

Sector\(n\)MeanSD
Private50326
Government50285
t / df / p3.62 / 98 / 0.0005

\(s_p = \sqrt{(49\cdot36 + 49\cdot25)/98} = 5.52\); \(t = 4/(5.52\times0.2) = 3.62\).

5. Result

\(t = 3.62,\ p = 0.0005 < 0.05\): \(H_0\) is rejected — perceived discrimination is significantly higher in the private sector in this illustrative sample.

Experiment 2 — Unemployment Duration and Mental Health

1. Problem

Does longer unemployment worsen mental health (PHQ-9 depression score)?

2. Aim

To compare PHQ-9 scores between recently and long-term unemployed groups and to measure the correlation between unemployment duration and PHQ-9.

3. Formula

\[ r = \frac{\sum(x-\bar x)(y-\bar y)}{\sqrt{\sum(x-\bar x)^2\sum(y-\bar y)^2}}, \qquad t_r = \frac{r\sqrt{n-2}}{\sqrt{1-r^2}} \]

Applying it:

  1. Questionnaire: employment status and duration; financial impact; validated PHQ-9 (0–27); coping strategies; demographics.
  2. Sampling: purposive — 50 recently unemployed (<6 months) + 50 long-term (≥6 months).
  3. Analysis: two-sample \(t\)-test on PHQ-9; Pearson correlation and simple regression of PHQ-9 on duration; chi-square on severity category × group.

4. Calculation

Blank working table:

AnalysisStatisticp
t-test (recent vs long-term)
Pearson r (duration vs PHQ-9)

Illustrative: recent \(\bar x = 8,\ s = 4,\ n = 50\); long-term \(\bar x = 13,\ s = 5,\ n = 50\); duration–PHQ-9 correlation \(r = 0.62\ (n = 50)\).

AnalysisStatisticp
t-test (recent vs long-term)t = 5.52 (df 98)< 0.001
Pearson r (duration vs PHQ-9)r = 0.62, t = 5.48< 0.001

5. Result

Long-term unemployed show significantly higher PHQ-9 scores (\(t = 5.52\)), and duration correlates moderately-strongly with depression (\(r = 0.62\)) — longer unemployment is associated with poorer mental health.

Experiment 3 — Impact of Subsidy Removal on Household Expenditure

1. Problem

Does subsidy removal raise fuel expenditure more for unorganised-sector workers than for the service class?

2. Aim

To compare the percentage increase in fuel expenditure between the two sectors and identify coping strategies.

3. Formula

\[ \%\,\text{increase} = \frac{\text{after} - \text{before}}{\text{before}}\times100, \qquad t = \frac{\bar x_1 - \bar x_2}{\sqrt{s_1^2/n_1 + s_2^2/n_2}} \]

Applying it:

  1. Questionnaire: sector, household size, income before/after; fuel expenditure before/after; coping-strategy checklist; a 5-point impact Likert item; demographics.
  2. Sampling: stratified — 75 service + 75 unorganised = 150.
  3. Analysis: percentage increase \(=\frac{\text{after}-\text{before}}{\text{before}}\times100\); two-sample \(t\)-test; chi-square (sector × coping strategy); Mann–Whitney U on the ordinal impact score.

4. Calculation

Blank working table:

Sector\(n\)Mean % increaseSD
Service class
Unorganised
t / p

Illustrative: service \(\bar x = 25\%,\ s = 8,\ n = 75\); unorganised \(\bar x = 40\%,\ s = 12,\ n = 75\).

Sector\(n\)Mean % increaseSD
Service class75258
Unorganised754012
t / p9.01 / < 0.001

\(t = (40-25)/\sqrt{8^2/75 + 12^2/75} = 15/1.665 = 9.01\).

5. Result

The unorganised sector faces a significantly larger fuel-expenditure increase (\(t = 9.01,\ p < 0.001\)) — subsidy removal is regressive, hitting informal workers hardest.

Experiment 4 — Online Shopping Behaviour Among College Students

1. Problem

Do male and female students differ in monthly online expenditure, and what drives spending?

2. Aim

To compare mean online expenditure by gender and model expenditure from pocket money, frequency and gender.

3. Formula

\[ t = \frac{\bar x_1 - \bar x_2}{\sqrt{s_1^2/n_1 + s_2^2/n_2}} \]

Applying it:

  1. Questionnaire: purchase frequency, average monthly spend, platform/payment preference; 8-item importance Likert scale; product-category checklist; demographics.
  2. Sampling: convenience — 50 male + 50 female = 100 students.
  3. Analysis: two-sample \(t\)-test on expenditure; chi-square (gender × category); multiple regression; factor analysis of the 8 importance items.

4. Calculation

Blank working table:

Gender\(n\)Mean spend (₹)SD
Male
Female
t / p

Illustrative: male \(\bar x = 1500,\ s = 600,\ n = 50\); female \(\bar x = 1800,\ s = 700,\ n = 50\).

Gender\(n\)Mean spend (₹)SD
Male501500600
Female501800700
t / p2.30 / 0.024

\(t = 300/\sqrt{600^2/50 + 700^2/50} = 300/130.4 = 2.30\) (df ≈ 96).

5. Result

Female students spend significantly more online (\(t = 2.30,\ p = 0.024 < 0.05\)) in this illustrative sample.

Experiment 5 — Customer Satisfaction with Public Transport

1. Problem

How satisfied are passengers with public bus transport, and which service aspects most drive overall satisfaction?

2. Aim

To assess the reliability of the satisfaction scale and identify the strongest predictors of overall satisfaction.

3. Formula

\[ \alpha = \frac{k}{k-1}\left(1 - \frac{\sum s_i^2}{s_T^2}\right) \]

Applying it:

  1. Questionnaire: usage pattern; a 10-item 5-point satisfaction scale (punctuality, cleanliness, comfort, safety, fare, coverage, frequency, driver behaviour, seat availability, overall); open suggestion; demographics.
  2. Sampling: systematic intercept — every 5th passenger at 5 stops, ~150 respondents.
  3. Analysis: Cronbach's \(\alpha\) for the scale; multiple regression (overall ~ 9 aspects); ANOVA across age/occupation groups.

4. Calculation

Blank working table:

QuantityValue
\(k\) (items)
\(\sum s_i^2\) / \(s_T^2\)
Cronbach's \(\alpha\)

Illustrative: \(k = 10\), sum of item variances \(\sum s_i^2 = 12\), total-score variance \(s_T^2 = 45\).

QuantityValue
\(k\)10
\(\sum s_i^2\) / \(s_T^2\)12 / 45
\(\alpha = \tfrac{10}{9}(1 - 12/45)\)0.815

5. Result

Cronbach's \(\alpha = 0.815\) (> 0.7) indicates good internal consistency, so the satisfaction scale is reliable and the regression on its components is meaningful.

Experiment 6 — Awareness and Usage of Digital Payment Systems

1. Problem

Does the adoption of digital payments differ across age groups?

2. Aim

To test the association between age group and digital-payment usage.

3. Formula

\[ \chi^2 = \sum \frac{(O-E)^2}{E}, \qquad df = (r-1)(c-1) \]

Applying it:

  1. Questionnaire: awareness checklist (UPI, cards, wallets, net banking); usage frequency and amount; 6-item barrier Likert scale; demographics (age, urban/rural).
  2. Sampling: quota — 50 each from ages 18–25, 26–40, 41–60, 60+ = 200.
  3. Analysis: chi-square (age group × user/non-user); one-way ANOVA on frequency; binary logistic regression for usage.

4. Calculation

Blank working table (users / non-users by age):

Age groupUsersNon-users
18–25
26–40
41–60
60+
χ² / df / p

Illustrative observed counts:

Age groupUsersNon-users
18–25455
26–404010
41–603020
60+1535
χ² / df / p46.15 / 3 / < 0.001

Expected users per group \(= 50\times130/200 = 32.5\); summing \((O-E)^2/E\) gives \(\chi^2 = 46.15\).

5. Result

\(\chi^2 = 46.15,\ df = 3,\ p < 0.001\): usage depends strongly on age — adoption falls sharply in older groups.

Experiment 7 — Study Habits and Academic Performance

1. Problem

How do study habits (hours, method, environment) relate to academic performance (GPA)?

2. Aim

To measure the correlation between study hours and GPA and identify the strongest predictors of GPA.

3. Formula

\[ t_r = \frac{r\sqrt{n-2}}{\sqrt{1-r^2}}, \qquad \widehat{\text{GPA}} = \beta_0 + \beta_1\,\text{Hours} + \beta_2\,\text{Sleep} + \beta_3\,\text{Attendance} \]

Applying it:

  1. Questionnaire: daily study hours, method, location, online-resource use; sleep, exercise, screen time; GPA, attendance; demographics.
  2. Sampling: simple random from the class register, ~80 students.
  3. Analysis: Pearson correlation (study hours vs GPA); multiple regression (GPA ~ study hours + sleep + attendance); \(t\)-test and ANOVA on method groups.

4. Calculation

Blank working table:

AnalysisStatisticp
Pearson r (hours vs GPA)

Illustrative: study hours vs GPA correlation \(r = 0.68\ (n = 80)\).

AnalysisStatisticp
Pearson r (hours vs GPA)r = 0.68, t = 8.19 (df 78)< 0.001

\(t_r = 0.68\sqrt{78}/\sqrt{1 - 0.68^2} = 8.19\).

5. Result

Study hours correlate strongly with GPA (\(r = 0.68,\ p < 0.001\)); \(r^2 = 0.46\), so about 46 % of GPA variation is explained by study time alone.

Experiment 8 — Food Consumption Patterns and Nutritional Awareness

1. Problem

Do urban and rural respondents differ in nutritional knowledge, and does knowledge relate to dietary quality?

2. Aim

To compare nutrition-quiz scores between urban and rural groups and correlate knowledge with dietary quality.

3. Formula

\[ t = \frac{\bar x_1 - \bar x_2}{\sqrt{s_1^2/n_1 + s_2^2/n_2}} \]

Applying it:

  1. Questionnaire: weekly consumption frequencies (fruit, vegetables, dairy, fast food, sugary drinks); a 10-item true/false nutrition quiz; a computed dietary-quality score; demographics.
  2. Sampling: stratified — 75 urban + 75 rural = 150.
  3. Analysis: two-sample \(t\)-test on knowledge; Pearson correlation (knowledge vs dietary quality); simple regression; chi-square (education × pass/fail).

4. Calculation

Blank working table:

Location\(n\)Mean knowledgeSD
Urban
Rural
t / p

Illustrative: urban \(\bar x = 7.2,\ s = 1.5,\ n = 75\); rural \(\bar x = 6.1,\ s = 1.8,\ n = 75\).

Location\(n\)Mean knowledgeSD
Urban757.21.5
Rural756.11.8
t / p4.07 / < 0.001

\(t = (7.2-6.1)/\sqrt{1.5^2/75 + 1.8^2/75} = 1.1/0.271 = 4.07\).

5. Result

Urban respondents score significantly higher on nutrition knowledge (\(t = 4.07,\ p < 0.001\)), consistent with better dietary quality where knowledge is higher.

Experiment 9 — Impact of Work-From-Home on Employee Productivity

1. Problem

Does work-from-home affect productivity and work-life balance, and do outcomes differ across industries?

2. Aim

To compare the mean work-life-balance score across industries (IT, education, finance) by one-way ANOVA.

3. Formula

\[ F = \frac{MS_{\text{between}}}{MS_{\text{within}}}, \qquad SS_{\text{between}} = \sum_g n_g(\bar x_g - \bar{\bar x})^2 \]

Applying it:

  1. Questionnaire: WFH status and duration; self-rated productivity change; a 5-item work-life-balance Likert scale; single-item job satisfaction; challenge checklist; demographics.
  2. Sampling: online survey, ~120 respondents across ≥3 industries.
  3. Analysis: chi-square (WFH status × productivity change); one-way ANOVA on work-life balance across industries; ordinal logistic regression; Cronbach's \(\alpha\).

4. Calculation

Blank working table:

Industry\(n\)Mean WLBSD
IT
Education
Finance
F / df / p

Illustrative: IT \(\bar x = 18,\ s = 4,\ n = 40\); Education \(\bar x = 20,\ s = 3.5,\ n = 40\); Finance \(\bar x = 16,\ s = 4.5,\ n = 40\).

Industry\(n\)Mean WLBSD
IT40184.0
Education40203.5
Finance40164.5
F / df / p9.90 / (2, 117) / < 0.001

\(\bar{\bar x} = 18\); \(SS_B = 40(0^2 + 2^2 + 2^2) = 320\), \(MS_B = 160\); \(SS_W = 39(16 + 12.25 + 20.25) = 1891.5\), \(MS_W = 16.17\); \(F = 160/16.17 = 9.90\).

5. Result

\(F = 9.90,\ p < 0.001\): work-life balance under WFH differs significantly across industries (highest for education, lowest for finance in this illustrative sample). A Tukey post-hoc test would identify which pairs differ.

Lab Record Format (to be followed for every experiment)

  1. 1. Problem — the research question.
  2. 2. Aim — the objective and hypothesis to be tested.
  3. 3. Formula — the formula, then the questionnaire design, sampling method and analysis plan.
  4. 4. Calculation — the analysis of the collected data (an illustrative worked example is shown; replace it with your own data).
  5. 5. Result — the inference, with interpretation in the research context.
Report structure for each project: title, abstract, introduction, methodology, results (with tables), conclusions, and the questionnaire as an appendix.
← Previous: Unit 5 Home →