ESM 304: Statistical Methods (2024) - Practical Lecture Notes
## 1. Normal Distribution
### Explanation:
The normal distribution is a bell-shaped curve describing how data is spread around the
mean. It's widely used because many natural processes follow this pattern.
### Example 1 (Excel):
Problem: A class has test scores: 60, 65, 70, 75, 80, 85, 90, 95, 100. Plot the normal
distribution.
Steps:
1. Enter the scores in column A.
2. Calculate the mean using =AVERAGE(A1:A9).
3. Calculate the standard deviation using =STDEV.S(A1:A9).
4. Create X values (60 to 100) in column B.
5. Use =NORM.DIST(B1, mean, std_dev, FALSE) in column C.
6. Plot B and C using Scatter Chart with smooth lines.
### Example 2 (SPSS):
Problem: Analyze the normality of income data.
Steps:
1. Enter or import income data.
2. Go to Graphs > Legacy Dialogs > Histogram.
3. Check "Display normal curve."
4. Click OK to see the bell shape.
### Assignment Questions:
1. Define normal distribution and state its properties.
2. Explain the empirical rule.
3. Plot a normal curve for 10 given values in Excel.
4. Interpret a histogram with a normal curve in SPSS.
5. Explain the significance of the mean and standard deviation.
### Quiz Questions:
1. What is the shape of a normal distribution?
2. What percent of data falls within ±1 SD?
3. Which function plots a normal curve in Excel?
4. What tool is used in SPSS to display a normal curve?
5. What does a Z-score represent?
## 2. Sampling Distribution
### Explanation:
A sampling distribution is the distribution of a sample statistic (like the mean) from
repeated samples taken from a population.
### Example 1 (Excel):
Problem: Generate 10 samples from a population of 100.
Steps:
1. In A2:A101, enter =RANDBETWEEN(50,100) to simulate population.
2. In B2:B6, enter =INDEX($A$2:$A$101, RANDBETWEEN(1,100)) for Sample 1.
3. Repeat in C, D... for more samples.
4. Below each sample, use =AVERAGE(B2:B6) to get sample mean.
5. Collect sample means and plot histogram.
### Example 2 (SPSS):
Problem: Estimate the sampling distribution of height.
Steps:
1. Enter height data for 100 people.
2. Use Select Cases > Random Sample (size = 5).
3. Analyze > Descriptive Statistics > Descriptives.
4. Save and record means. Repeat for 10 samples.
### Assignment Questions:
1. Define sampling distribution.
2. What is the standard error?
3. Simulate 3 samples of size 4 in Excel.
4. What does the central limit theorem say?
5. What affects the shape of the sampling distribution?
### Quiz Questions:
1. Sampling distribution involves repeated _____?
2. What Excel function can be used to pull a random sample?
3. What happens to SE as sample size increases?
4. Which SPSS feature allows random sampling?
5. Sampling distribution helps in making _____ about population?
## 3. Estimation
### Explanation:
Estimation is used to infer the value of a population parameter from a sample statistic. Two
types are point estimation and interval estimation.
### Example 1 (Excel):
Problem: Estimate the population mean from a sample.
Steps:
1. Input sample data in column A.
2. Use =AVERAGE(A1:A10) for point estimate of mean.
3. Use =CONFIDENCE.NORM(alpha, STDEV.S(A1:A10), COUNT(A1:A10)) for margin of error.
4. Compute confidence interval: Mean ± Margin of Error.
### Example 2 (SPSS):
Problem: Estimate the mean income with 95% CI.
Steps:
1. Analyze > Descriptive Statistics > Explore.
2. Put income under Dependent List.
3. Click Statistics > Confidence Interval for Mean (default 95%).
4. View confidence interval in the output.
### Assignment Questions:
1. Define point and interval estimation.
2. How is confidence level related to margin of error?
3. Estimate population mean from a given sample in Excel.
4. What does a 95% confidence interval mean?
5. Use SPSS to generate a confidence interval.
### Quiz Questions:
1. What is a point estimate?
2. What is an interval estimate?
3. What function gives confidence margin in Excel?
4. What is the default CI in SPSS?
5. Larger sample sizes make intervals _____?
## 4. Student’s t-Distribution
### Explanation:
The t-distribution is used instead of normal distribution when the sample size is small (n <
30) and population standard deviation is unknown.
### Example 1 (Excel):
Problem: Calculate the t-critical value.
Steps:
1. Sample size = 10, df = 9
2. Use =T.INV.2T(0.05, 9) for two-tailed test.
### Example 2 (SPSS):
Problem: Conduct one-sample t-test.
Steps:
1. Analyze > Compare Means > One-Sample T-Test.
2. Enter test variable and test value.
3. SPSS will output t, df, and p-value.
### Assignment Questions:
1. Define t-distribution.
2. When is it used over normal distribution?
3. Calculate t-critical value in Excel.
4. Conduct one-sample t-test in SPSS.
5. Interpret p-value in t-test.
### Quiz Questions:
1. What shape is t-distribution?
2. When do you use t instead of z?
3. Excel formula for two-tailed t-value?
4. In SPSS, which menu for t-test?
5. What is df in t-test?
## 5. Statistical Decision Theory
### Explanation:
Statistical decision theory involves making decisions under uncertainty using probability
models and decision criteria.
### Example 1 (Excel):
Problem: Decide whether to invest in Project A or B based on expected payoff.
Steps:
1. Enter possible states (e.g., Good, Fair, Poor) in A2:A4.
2. Enter payoffs for each project in B2:C4.
3. Enter probabilities in D2:D4.
4. Compute expected value: =SUMPRODUCT(B2:B4, D2:D4) and =SUMPRODUCT(C2:C4,
D2:D4).
5. Choose project with higher expected value.
### Example 2 (SPSS):
Problem: Analyze outcomes of multiple decisions.
Steps:
1. Enter states and payoffs in columns.
2. Use Transform > Compute Variable to calculate expected value.
3. Compare computed outcomes.
### Assignment Questions:
1. Define decision theory.
2. What is expected value?
3. How do you choose between two uncertain projects in Excel?
4. What does SPSS compute in decision modeling?
5. What is the role of probabilities in decisions?
### Quiz Questions:
1. What Excel function computes expected value?
2. What is the first step in decision analysis?
3. What is a payoff matrix?
4. Which tool in SPSS can compute value functions?
5. Decisions under uncertainty are based on ____?
## 6. Hypothesis Testing
### Explanation:
Hypothesis testing helps determine whether a statement about a population is true using
sample data.
### Example 1 (Excel):
Problem: Test if the average score is 70.
Steps:
1. Enter scores in column A.
2. Use =T.TEST(A1:A10, 70, 2, 1) to test mean = 70.
3. Interpret p-value: if p < 0.05, reject H0.
### Example 2 (SPSS):
Problem: One-sample t-test for mean = 70.
Steps:
1. Analyze > Compare Means > One-Sample T-Test.
2. Enter test value (70).
3. Check Sig. (2-tailed) for result.
### Assignment Questions:
1. Define null and alternative hypotheses.
2. What does p-value represent?
3. Use Excel to test a mean hypothesis.
4. Use SPSS to test a proportion hypothesis.
5. Define significance level.
### Quiz Questions:
1. What is H0?
2. When is H0 rejected?
3. Excel formula for t-test?
4. Where is p-value in SPSS output?
5. Typical alpha value?
## 7. Chi-square Test
### Explanation:
Chi-square test checks for independence or goodness of fit between variables.
### Example 1 (Excel):
Problem: Test if two attributes are independent.
Steps:
1. Create observed frequency table.
2. Create expected frequency table.
3. Use formula: =(Observed - Expected)^2 / Expected.
4. Sum all to get Chi-square stat.
5. Compare with critical value.
### Example 2 (SPSS):
Problem: Test gender vs product choice.
Steps:
1. Analyze > Descriptive Statistics > Crosstabs.
2. Select variables, click Statistics > Chi-square.
3. Click OK and interpret Chi-square stat and p-value.
### Assignment Questions:
1. Define Chi-square test.
2. How are expected frequencies computed?
3. Use Excel to test association between gender and preference.
4. Use SPSS for Chi-square test.
5. Interpret a p-value of 0.03.
### Quiz Questions:
1. What does Chi-square test check?
2. Which Excel formula gives Chi-square stat?
3. SPSS menu for Chi-square?
4. What is the test statistic formula?
5. What does high Chi-square value indicate?
## 8. Goodness of Fit
### Explanation:
This test determines how well observed data matches expected distribution.
### Example 1 (Excel):
Problem: Test if dice is fair.
Steps:
1. Enter observed frequencies (1-6 outcomes).
2. Enter expected (e.g., 10 per outcome if total = 60).
3. Use =(O-E)^2/E formula for each category.
4. Sum results for Chi-square statistic.
### Example 2 (SPSS):
Problem: Test normality of scores.
Steps:
1. Analyze > Descriptive > Frequencies.
2. Click Charts > Normal Curve.
3. Use Analyze > Nonparametric > Chi-square.
### Assignment Questions:
1. Define goodness of fit.
2. What is the null hypothesis?
3. Compute test statistic in Excel.
4. Use SPSS to test data fit to normal distribution.
5. Interpret Chi-square results.
### Quiz Questions:
1. What does "goodness of fit" mean?
2. What distribution is typically tested?
3. Excel formula for fit test?
4. Where in SPSS to run Chi-square goodness of fit?
5. What does low p-value indicate?
## 9. Linear Regression
### Explanation:
Linear regression predicts a dependent variable based on one or more independent
variables.
### Example 1 (Excel):
Problem: Predict sales based on ads.
Steps:
1. Enter ad spend in A, sales in B.
2. Use =LINEST(B1:B10, A1:A10, TRUE, TRUE).
3. Use equation y = a + bx to predict.
### Example 2 (SPSS):
Problem: Regress income on education.
Steps:
1. Analyze > Regression > Linear.
2. Put income as dependent, education as independent.
3. Interpret coefficients and R².
### Assignment Questions:
1. What is regression used for?
2. What is R²?
3. Perform regression in Excel.
4. What does a coefficient mean?
5. Use SPSS for regression.
### Quiz Questions:
1. What’s the formula for simple linear regression?
2. What does R² show?
3. What function in Excel runs regression?
4. SPSS regression menu?
5. Slope indicates ____?
## 10. Correlation
### Explanation:
Correlation measures the strength and direction of linear relationship between two
variables.
### Example 1 (Excel):
Problem: Test correlation between height and weight.
Steps:
1. Enter height in A, weight in B.
2. Use =CORREL(A1:A10, B1:B10).
### Example 2 (SPSS):
Steps:
1. Analyze > Correlate > Bivariate.
2. Select variables and check Pearson.
3. Click OK and interpret result.
### Assignment Questions:
1. Define correlation.
2. What are the types?
3. Compute correlation in Excel.
4. Use SPSS to interpret correlation matrix.
5. What does r = -0.8 mean?
### Quiz Questions:
1. What does correlation measure?
2. What’s max and min value?
3. Excel correlation function?
4. SPSS path for Pearson correlation?
5. r = 0 means?
## 11. Index Numbers
### Explanation:
Index numbers compare changes in economic variables over time.
### Example 1 (Excel):
Problem: Construct price index.
Steps:
1. Base year prices in A, current prices in B.
2. Use =B1/A1*100 to get price index.
### Example 2 (SPSS):
1. Enter base and current prices.
2. Compute index using Transform > Compute.
### Assignment Questions:
1. What is a price index?
2. How to calculate Laspeyres index?
3. Create index in Excel.
4. Use SPSS to compute index.
5. What does index >100 mean?
### Quiz Questions:
1. What is base year?
2. What does index = 120 mean?
3. Excel formula for index?
4. SPSS compute path?
5. Index compares values over ____?
## 12. Time Series Analysis
### Explanation:
Time series analysis involves analyzing data over time to identify trends, cycles, or
seasonality.
### Example 1 (Excel):
Problem: Analyze sales over 12 months.
Steps:
1. Enter months in A, sales in B.
2. Insert Line Chart.
3. Use =AVERAGE(B1:B3) in C to compute moving average.
### Example 2 (SPSS):
1. Analyze > Forecasting > Create Models.
2. Choose time series variable.
3. Apply Moving Average or ARIMA.
### Assignment Questions:
1. What is time series?
2. What is moving average?
3. Analyze sales trend in Excel.
4. Use SPSS for forecast.
5. What is seasonality?
### Quiz Questions:
1. What are time series components?
2. Which chart shows trend?
3. Formula for moving average in Excel?
4. SPSS menu for time series?
5. Trend reflects ____ change?
