A very special place

Santa Teresa, a bucolic neighborhood in Rio de Janeiro, Brazil, is a very special place to me. It was there that I first lived after moving from my parents’ house. And I picked that place because it seemed to me like an oasis in the middle of the city’s concrete jungle. Santa Teresa is on top of hill and at many places of that neighborhood one can have a spectacular view of Rio de Janeiro. The one that I like most is granted on the windows of Valentin’s Castle, where it is possible to see Rio de Janeiro downtown and the round banks of Guanabara Bay crossed by the sinuous line of Rio-Niteroi bridge.

Graph Language

The graph below shows 18-year time series of Iranian GDP growth rate and the oil barrel price. The oil world crisis of 1986 sunk oil barrel prices, but an impressive recovery can be seen in the subsequent four years, followed by a two-year period of stabilization. In 1992, however, the oil prices fell again, managing to rise moderately until 1997 when dip before a new short period of stagnation.

Iranian economy is intrinsically dependent on the international oil trade and the graph depicts this close relationship. Deep falls and skyrocketing rises on the international oil barrel price are followed with similar trends in Iranian GDP growth rate. Even a quick dip in the oil prices as observed in years 2000/2001, is reflected in the Iranian GDP growth rate.

The graph below depicts the changes in the state tax revenue between 1999 and 2009. From 2000 to the first two quarters of 2002, state tax revenue decreased sharply when began a choppy uprising movement until the third quarter of 2005 when achieved the peak of the whole period. It is worth mentioning the rapid growth during the second half of 2002 and during the first semester of 2005. But starting in the 2005 peak, tax revenue decreased steadily up to first quarter of 2009, when it dropped abruptly again.

All these variations and especially the intense drops observed on the graphs make it easier to understand why states were under financial risks during this decade. Without consistent tax revenue it was hard to state officials to define public policies priorities and finance them appropriately.

Hedging in Data Commentaries

Dear Professor Burn,

In attention to your request, I prepared the table below to highlight general differences between the regular and makeup exams.

	Regular Exam	Makeup exam
Average score (out of 100)	86	71
Time administered	Wednesday, 7:00 pm	Friday, 4:00pm
Difficulty questions	Average	Average
# of students	125	25
Proctor	Professor	Teaching assistant
Board examples	Yes	No (not considered necessary)
Room environment	About 20C	About 28C

The average score of the makeup exam was 15 points below the average score of the regular exam. Only 25 students attended the makeup exam and the room temperature was a bit higher during the second exam.

I don’t believe there is necessarily any particular reason for lower average performance in the makeup exam. The level of questions difficulty was pretty much the same in the two exams.

One point, however, that we could consider is the number of students who took the makeup exam. There was only 25 students during the second exam and if even very few of them performed poorly in the test that would be enough to substantially decrease the average result of the group. I suggest that we track the worst students’ grades to see if they are concentrated on the makeup exam.

It would also be useful to check if board examples provided during the regular exam helped students score better. Finally, it would be helpful to ask the students if they felt uncomfortable with room temperature during the makeup test. Some of them might have felt disturbed by the environmental conditions.

Best regards,

John