Absolutely! Let's go step by step through question 3.14, which deals with **ANOVA (Analysis of Variance)** for testing the effect of subgrade soil types on topsoil moisture content. --- ### **Given Data:** - 4 types of subgrade soil - 5 samples for each type (so, total samples = 4 × 5 = 20) - Total Sum of Squares (SST) = 280 - Sum of Squares Among types (SSA) = 120 --- ### **a. Set up an analysis of variance table for these results.** The typical ANOVA table: | Source | Degrees of Freedom | Sum of Squares | Mean Square | F-ratio | |-------------|-------------------|----------------|-------------|----------| | Between Groups (SSA) | k-1 | SSA | SSA/(k-1) | MS(A)/MS(E) | | Within Groups (SSE) | N-k | SSE | SSE/(N-k) | | | Total | N-1 | SST | | | Where: - \( k \) = number of groups = **4** - \( n \) = samples per group = **5** - \( N \) = total samples = **20** - \( SST \) = total sum of squares = **280** - \( SSA \) = sum of squares among = **120** - \( SSE \) = sum of squares error = \( SST - SSA = 280 - 120 = 160 \) **Degrees of Freedom:** - Between: \( k-1 = 3 \) - Within: \( N-k = 16 \) - Total: \( N-1 = 19 \) **Mean Squares:** - Between: \( MS_A = SSA/(k-1) = 120/3 = 40 \) - Within: \( MS_E = SSE/(N-k) = 160/16 = 10 \) **F-ratio:** \( F = MS_A / MS_E = 40/10 = 4 \) So, the table: | Source | df | SS | MS | F | |------------------|-----|------|------|--------| | Among (Between) | 3 | 120 | 40 | 4 | | Error (Within) | 16 | 160 | 10 | | | Total | 19 | 280 | | | --- ### **b. Set up a mathematical model** This is a **one-way ANOVA model**: \[ Y_{ij} = \mu + \tau_i + \epsilon_{ij} \] Where: - \( Y_{ij} \): moisture content for the j-th sample of the i-th soil type - \( \mu \): overall mean - \( \tau_i \): effect of the i-th soil type (\( i = 1,2,3,4 \)) - \( \epsilon_{ij} \): random error (assumed i.i.d. normal, mean 0, variance \(\sigma^2\)) **Assumptions:** - Errors (\( \epsilon_{ij} \)) are independent and normally distributed with mean 0 and constant variance. - The samples are randomly selected. - The effects (\( \tau_i \)) are fixed. --- ### **c. Set up a test of the hypothesis that the four types of subgrade soil have the same effect** **Null Hypothesis (\( H_0 \)):** \[ \tau_1 = \tau_2 = \tau_3 = \tau_4 = 0 \] (So, all means equal: \( \mu_1 = \mu_2 = \mu_3 = \mu_4 \)) **Alternative Hypothesis (\( H_A \)):** At least one \( \tau_i \neq 0 \) (at least one mean is different). Use the F-test from the ANOVA table: \[ F = \frac{MS_A}{MS_E} \] Compare computed F (here, 4) to the critical value \( F_{3,16,\alpha} \) from F-tables (with 3 and 16 df). --- ### **d. Set up a set of orthogonal contrasts for this problem** - With 4 groups, you can have 3 independent (orthogonal) contrasts. One possible set (assuming groups are 1,2,3,4): 1. Compare group 1 vs others: \[ C_1: \frac{3}{4}\mu_1 - \frac{1}{4}(\mu_2+\mu_3+\mu_4) \] 2. Compare group 2 vs 3,4: \[ C_2: \frac{1}{2}\mu_2 - \frac{1}{4}(\mu_3+\mu_4) \] 3. Compare group 3 vs 4: \[ C_3: \frac{1}{2}\mu_3 - \frac{1}{2}\mu_4 \] Make sure coefficients sum to zero and contrasts are orthogonal (dot products = 0). --- ### **e. Explain briefly how to set up a test on means after the ANOVA** If ANOVA is significant (rejects \( H_0 \)), use **multiple comparison procedures** (like Tukey's, Bonferroni, LSD, etc.) to test which means differ: - Compute the difference between group means. - Calculate standard error for the difference (using \( MS_E \)). - Use the appropriate critical value (from t or q tables) to determine if the difference is significant. --- ### **f. Set up an expression for 90% confidence limits on the mean of type 2 subgrade soil** The confidence interval for the mean of group 2: \[ \bar{Y}_{2\cdot} \pm t_{0.05,16} \times \sqrt{\frac{MS_E}{n}} \] Where: - \( \bar{Y}_{2\cdot} \): sample mean for type 2 - \( t_{0.05,16} \): t-value for 90% CI, 16 df (from t-table; approx 1.746) - \( MS_E = 10 \) - \( n = 5 \) So, \[ \bar{Y}_{2\cdot} \pm 1.746 \times \sqrt{\frac{10}{5}} = \bar{Y}_{2\cdot} \pm 1.746 \times \sqrt{2} = \bar{Y}_{2\cdot} \pm 1.746 \times 1.414 = \bar{Y}_{2\cdot} \pm 2.47 \] --- **If you need a deeper explanation or examples for any part, let me know!**

Certainly! Here's a concise summary: 1. ANOVA table: df=3,16; SSA=120, SSE=160, MS_A=40, MS_E=10, F=4. 2. Model: \(Y_{ij} = \mu + \tau_i + \epsilon_{ij}\), with errors normal, independent, equal variance. 3. Test \(H_0: \tau_1=\tau_2=\tau_3=\tau_4=0\) using F=4; compare to critical F. 4. Orthogonal contrasts: e.g., group 1 vs. others, group 2 vs. 3 & 4, group 3 vs. 4. 5. 90% CI for type 2 mean: \(\bar{Y}_2 \pm 1.746 \times \sqrt{10/5} \approx \bar{Y}_2 \pm 2.47\).