Republic of the Philippines

CEBU TECHNOLOGICAL UNIVERSITY

MAIN CAMPUS
M. J. Cuenco Avenue Cor. R. Palma Street, Cebu City, Philippines
Website: http://www.ctu.edu.ph E-mail: [email protected]
Phone: +6332 402 4060 loc. 1109

COLLEGE OF ARTS AND SCIENCES

Name: Shylen Kate Y. Abayan Date: October 17, 2024

Course/Year/Section: BSSTAT 3-1 Econometrics Activity 3

1. A. Test the hypothesis that B2 = 10.5 against the alternative that B2 > 10.5 at 0.01 level of
significance.
B2 = 10.5, B2 > 10.5
t = 0.142099 - 10.5 / 0.00196623
t = -10.357901 / 0.00196623
t = -5,267.90

Interpretation:
Given that the t-statistic (-5267.9) is significantly lower than the critical value (2.40), we fail to
reject the null hypothesis. The p-value of 1.000 suggests that there is no evidence to support the claim
that β2 > 10.5.

B. Test the hypothesis that B2 = 6.5 against the alternative that B2 < 6.5 at 0.05 level of significance.
B2 = 6.5, B2 < 6.5
t = (b2 - 6.5)/se(b2)
t = 0.142099 - 6.5/0.00196623
t = -6.357901/0.00196623
t = -3,233.55

Intepretation:
Since the t-statistic (-3233.55) is lower than the critical value (-1.68), we reject the null
hypothesis. The p-value is nearly zero (0.000), providing strong evidence that β2 < 6.5.

C. Test the hypothesis that B2 = 8.5 against the alternative that B2 ≠ 8.5 at 0.05 level of significance.
B2 = 8.5, B2 ≠ 8.5
t = (B2 − 8.5)/se(b2)
t = -8.357901/0.00196623
t = -4,250.72

Intepretation:
Since the t-statistic (-4250.72) is significantly beyond the range of -2.010 to 2.010, we reject the
null hypothesis. The p-value is essentially zero, providing compelling evidence that β2 ≠ 8.5.
2. A. Using the program Gretl, perform an analysis of variance forqualifying pscore variable depending
on the small, which indicates whether the class is small or standard.

B. Test that the average rating is independent of class size, measured according to the variable small.
F(1, 4324) = 11639.7 / 741.808 = 15.691
P-value: 7.58e-005

Interpretation:
 F value (15.691): The variance in class sizes is greater than within each group, which means that
class size significantly affects the average rating.
 P value (7.58e-005): Average rating is dependent on the class size.

The very small p-value indicates that the treatment's effect on pscore is significant and not merely a
random occurrence (using a standard level like 0.05). This suggests that the average scores of the two
groups differ. The analysis reveals a distinct difference in pscore between the groups: the with
treatment group (small = 1) has a higher average score of 54.93, while without the treatment group
(small = 0) has an average score of 51.59. The large F-statistic (15.691) and low p-value (7.58e-005)
support the conclusion that this difference is real rather than coincidental.

C. Perform a least square regression of the rating pscore on the size.

D. What is your conclusion about the effect of class size on academic performance?
 Class size seems to significantly influence academic performance. Smaller classes are linked to
higher average scores, indicating that decreasing class size could enhance student outcomes.

E. How would your conclusion if during the experiment students had changed from class originally
assigned to other of different sizes?
 If students switched classes during the experiment, particularly due to parental influence, it
could lead to unfair and biased results. The random way classes would be compromised.
Students who changed classes might have characteristics that influence their academic
performance, such as increased motivation. This would complicate the ability to attribute
differences in test scores solely to class size.

3. A. ln(V1) = 3,67 + 0,48 ln(V2);

Interpretation:
The findings indicate that a 1% increase in total annual spending results in an approximate
0.48% rise in food expenditure. This demonstrates a clear relationship between overall family spending
and food costs. The very low p-values confirm that this correlation is statistically significant, affirming
that total spending indeed influences food spending.

Additionally, the model's R-squared value reveals that around 26.8% of the variability in food
spending can be attributed to changes in total spending. While this percentage isn't particularly high, it
still suggests that total spending has some effect on food expenditures. The Adjusted R-squared being
similar to the R-squared indicates the model's reliability remains stable with a single predictor.

In conclusion, the results clearly illustrate that total family spending significantly affects annual
food spending.

B. V1 = −164567 + 2163 ln(V2);

Interpretation:
The findings indicate that a 1% increase in total annual expenditure corresponds to an
approximate rise of 2163.14 euros in family food spending. This suggests a clear correlation between
overall family spending and food expenses. The extremely low p-values (well below 0.05) confirm the
statistical significance of these results, highlighting that total spending is a key determinant of food
costs.

Additionally, the R-squared value reveals that about 26.9% of the variability in food spending
can be attributed to changes in overall spending. While this percentage may not be particularly high, it
still indicates that total expenditure influences food costs to some degree. The Adjusted R-squared value
is closely aligned with the R-squared, suggesting that the model's accuracy remains consistent when
accounting for the predictors.

In summary, these results imply that as families increase their overall spending, there is a
corresponding tendency for food expenditures to rise, and this relationship is statistically significant.

C. (V1/V2) × 100 = 156,89 − 13,32 ln(V2) ( “ Working-Lesser ” specification, which consider the
determinants of food expenditure as a percentage of total expenditure).

Interpretation:
The analysis reveals that a 1% increase in a family's total spending (V2) is associated with
approximately a 12.62% decrease in the percentage of their budget allocated to food. This indicates that
families with higher overall spending tend to devote a smaller fraction of their budget to food expenses.
Additionally, when the wife is employed (V10 = 1), the percentage spent on food is about 3.70% lower
than when she is not working. Thus, families with a working wife also spend a smaller share of their total
budget on food.

The low p-values (well below 0.05) confirm that these results are statistically significant,
indicating that both total spending and the wife's employment status have a clear impact on food
expenditure. The R-squared value of 33.37% suggests that about one-third of the variability in the
percentage spent on food can be attributed to these two factors, implying that there are likely
additional influences.

Overall, the findings indicate that families with higher total spending and those with working
wives allocate a smaller portion of their budget to food.