Quantitative Methods-I
Assignment
Sem-I
Q1) The University wishes to determine whether the observed increase in preference for
Chinese food from 30% to 40% is statistically significant. This requires hypothesis testing
using a z-test for proportions.
Let p (Historical proportion) =0.30
And p^ (Sample proportion) = 0.40
n=100
1. Hypothesis Formulation-
Null Hypothesis (H0) = 0.30
Alternate Hypothesis (H1) ≠ 0.30
2. Calculating Standard Error-
Standard Error= √[p(1-p)/n]
= √[(0.3x0.7)/100]
= 0.0458
3. Calculating z-score-
Z= (p^- p)/SE
= (0.40-0.30)/0.0458
= 2.183
4. Interpretation-
To come to a decision, the attained z-score is compared against a critical value.
Assuming a significance level of 5%, Z-critical would be + 1.96.
Therefore, since 2.18>1.96, the null hypothesis will be rejected. This shows that the
preference for Chinese food has changed from the historical 30%
Thus, the increase is statistically significant. The university should adapt their services
accordingly.
Q2(A) Different types of probability offers-
Marginal Probability: Shows overall gender mix and overall purchase frequency. It is
easy to compute and interpret, but it ignores any relationship between gender and
buying behavior. It’s a good starting point for budgeting and broad audience sizing. It
gives a vast view of the customer base and helps in inventory planning. But it lacks
depth. It treats the entire segment as a single unit which leads to generic marketing.
Joint Probability: It describes how often specific gender-purchase-frequency
combinations occur. It reveals the data structure and identifies prominent segments,
but does not explain whether gender drives the behavior of purchasing. It is essential
for discovering interaction patterns.
Conditional Probability: It measures the likelihood of a purchasing pattern given a
gender or vice-versa. It directly informs the segment about offers through specific
messaging, enabling a more precise targeting. While it is the best method for
targeting, it requires more specific data and can be misleading if the sample size for a
specific sub-segment is too small.
Independence vs Dependence:
The retail chain is debating whether gender and purchasing frequency are independent.
If Independent: This means gender has zero influence on how much someone buys. If
the chain assumes independence, they would use a one-size-fits-all campaign. This is
cheaper but often results in wasted ad spent on disinterested consumers.
If Dependent: One variable provides information about the other. If data shows that
men tend to buy more frequently than women or vice-versa, the events are dependent.
The retail chain should assess dependence between gender and purchasing behavior with an
appropriate statistical test and evaluate effect size. If dependence is statistically significant
and practically meaningful, proceed with conditional-probability-based segmentation. Then
the chain should build segment-specific models using conditional probabilities (eg,
probability of high purchase frequency given gender). The chain should use a hybrid
approach where they rely on marginal probabilities for high-level budgeting and market
sizing but use conditional probability-based strategies for the top segments where data
supports reliable estimates. This will ensure campaign effectiveness.
Q2(B) Mean(u)= 150,000
Standard Deviation= 20,000
Target Value (X)= 180,000
Step 1 Calculating the Z-score
The Z-score standardizes the target value by determining how many standard
deviations it lies from the mean:
z= (X-u)/SD
= (180,000-150,000)/20,000
= 1.5
Step 2 Find the Probability
Using the standard normal distribution table, the area to the left of z=1.5
(probability of sales< 180,000) is 0.9332. To find the probability of sales
exceeding 180,000, we subtract it from 1:
P(X> 180,000) = 1-0.9332= 0.0668
Thus, the probability that a randomly selected store earns more than $180,000 is 6.68%
From a managerial perspective, this result indicates that premium performance is a rare event
under the current operational model. Only about 17 (250 x 0.0668) out of 250 stores are
expected to meet or exceed this threshold in any given month. The result suggests that a store
hitting this target is performing significiantly better than the average store.
The Premium store threshold of $180,000 appears to be strict but reasonable, depending on
the intended purpose of the classification. If the premium label is intended to be a prestigious
designation for top-tier performers, a 6.68% attainment rate is appropriate. It ensures the
status is not diluted by being too easy to achieve.
If the management intended for this to be a standard motivational target for the majority of
stores, it is likely too high. Since over 93% of the stores are currently failing to hit this mark,
it may lead to employee burnout or demotivation if the gap between average performance and
the reward threshold feels insurmountable.