Homework 5: Exact and Approximate Inference for Call Center Waiting Times
Background
Waiting times in service systems are often modeled using the exponential distribution.
Suppose
\[ X_1,\dots,X_n \sim \Exp(\theta), \]
where \(\theta\) is the mean waiting time.
For exponential data, an important fact is that
\[ \frac{2\sum_{i=1}^n X_i}{\theta} \sim \chi^2_{2n}. \]
This gives an exact method for inference about \(\theta\).
We can also use a Normal approximation based on the Central Limit Theorem.
Data
The context for this assignment is based on real data from the New York City Open Data portal:
NYC Open Data. 311 Service Requests from 2010 to Present.
The observations used here are inter-arrival waiting times (in seconds) computed from service-request timestamps for a fixed complaint type.
For this homework, we use a subset of 20 waiting times selected from that real dataset.
The following waiting times (in seconds) were observed:
4 7 8 11 29 34 36 41 47 49
53 65 79 89 134 134 165 178 227 375
Thus,
\[ n = 20. \]
Question
Suppose that a service target states that the mean waiting time should not exceed 60 seconds.
We test
\[ H_0:\theta = 60 \qquad\text{vs}\qquad H_a:\theta > 60. \]
Use significance level
\[ \alpha = 0.05. \]
Part 1: Looking at the Data
Compute the sample mean \(\barX\).
Make a simple plot of the data, such as a histogram, dotplot, or stem-and-leaf display.
Comment briefly on whether the data appear right-skewed.
Explain briefly why an exponential distribution might be a reasonable model for waiting times.
Part 2: Exact Hypothesis Test
For the exact test, use
\[ \frac{2\sum_{i=1}^n X_i}{\theta} \sim \chi^2_{2n}. \]
Compute the test statistic for testing \(H_0:\theta=60\).
Compute the p-value.
State your conclusion in context.
Part 3: Exact Upper Confidence Bound
Using the exponential model, construct a 95% upper confidence bound for \(\theta\) of the form
\[ [0,\theta_U]. \]
Compute this bound and interpret it in context.
Part 4: Normal Approximation
Using the Central Limit Theorem, approximate the distribution of \(\barX\) by a Normal distribution and construct the corresponding test statistic.
Compute the approximate p-value.
Construct the approximate 95% upper confidence bound.
Part 5: Comparison and Discussion
Compare the exact exponential inference and the Normal approximation.
Discuss:
the two p-values,
the two upper confidence bounds,
whether the two methods lead to the same practical conclusion,
whether the Normal approximation seems reasonable here.
Submission
Your work should include:
- numerical calculations,
- a plot of the data,
- clear conclusions in context,
- and enough explanation that another student could follow your reasoning.