Question

4. You are interested in analyzing whether pollution affects productivity at call centers. You have a dataset that lists daily calls per worker made from call centers in different cities. The data also shows the Air Quality Index at the city on that date. The Air Quality Index is a measure of the number of particles in the air that ranges from 0 to 500. Lower levels indicate less pollution. The following lines are an extract from the full dataset: call center id day of the week calls_per_ worker AOI 100 02/01/2010 Monday 02/02/2010 Tuesday 02/01/2010 Monday 02/02/2010 Tuesday 02/01/2010 Monday 02/02/2010 Tuesday 02/01/2010 Monday 02/02/2010 Tuesday 81 75 85 100 100 70 a) b) c) What type of dataset is this (cross-section, time series, panel)? What is the unit of observation? Remember, cach row in a dataset is a unit of observation. Describe why it is relevant to use calls per worker (as opposed to total calls) to measure productivity at each call center. [Hint: think about the number of calls that occur at a call center with 10 workers and one with 100.] Write the Stata command (or commands) you would use to get the median calls per worker on Mondays. You regress calls per worker on Air Quality Index and get the results shown below: d) Source MS 1,050 367.10 0.0000 -0.2594 Adj R-squared0.2587 12.562 Number of obs F (, 1048) Model57928.7844 Residual 165377.178 1 57928.7844 Prob >E 1,048 157.802651 R-squared Total223305.963 1,049 212.875084 Root MSE calls Coef Std. Err (95% conf. Interval] AQI 2012311 0105028 19.16 0.000 1806222 22184 53.79149.750834 71.64 0.000 52.31818 55.26479 e Write the resulting regression equation. Be precise. t) In a sentence, interpret the coefficient for AQI in the regression. g) What are the predicted calls per worker for a call center in a city with an AQI of 200? h) What is the predicted change in calls per worker for a call center in a city where AQI decreases from 100 to 50?

Answer 1

a. Since the above data set pertains to the single time period and not measuring the calls per workers across the years, thus, data set is cross sectional data set in the above case.

b. For the X variable, the unit of measurement is number of calls and for the Y variable which is the Air Quality Index, there is no unit of measurement because it is an index value.

c. To measure productivity, it is important to use calls per worker because it tells the number of calls being attended by each person and thus is a better measure of productivity as compared to total number of calls.

d. The median calls per worker on Monday = Total number of calls on Monday/ Total number of workers working on Monday