EC 41, UCLA Winter 2009
Name (print)________________________________
Midterm #1
– 1/26/09
TA: Name__________________________ & Section Time_____________
 The normal table and useful formulas are on the last page of this exam.
 Only pens, pencils and erasers may be used, this is a closed book, closed note, exam.
 Students may use a calculator, but nothing that can access the internet.
 Write noninteger answers to 3 significant digits, e.g., 333, or 3.33 or .0333
 This exam consists of 10 True/False (20 points), 10 short answer (30 points) and 5 longer questions (50 points)
 Clearly write answers on this exam.
No points are awarded for illegible answers.
 Be prepared to show a photo ID during the exam (e.g., UCLA ID)
 You may leave when finished.
Do not disrupt those still taking the exam.
I. Circle T for True or F for False (2 points each)
1) T or F
A
categorical variable
takes numerical values for which arithmetic operations have meaning.
2) T or F
In the demand for murder example, time series data for the U.S. suggested that increased use of capital
punishment would reduce murders per capita.
3) T or F
Data from a back to back stemplots such as incomes for unrelated American and Japanese individuals can be
used to create a meaningful scatter plot.
4) T or F
Data on price, income, or wealth are typically skewed leftward (downward) toward smaller values.
5) T or F
In a standard OLS regression of Y on X, we assume X is measured without error but that there may be errors in
the measurement of Y.
6) T or F
If X and Y have the same standard deviations, s
X
= s
Y
, then it will be impossible for either the b
1
estimated by
regression of Y on X, or the b
1
’ estimated by the reverse regression of X on Y, to have a value greater than 1.
7) T or F
In a standard linear regression of Y on X, the squared sum of the horizontal distances between the estimated
line and the actual X values is minimized.
8) T or F
For the regression of Y on X if
X is two standard deviations above
X
then the corresponding predicted value
of Y (
ˆ
Y
) must be more than two standard deviations above
Y
(assume the correlation between X and Y is not 1 or 1).
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
This is the end of the preview.
Sign up
to
access the rest of the document.
 Fall '07
 Guggenberger
 Linear Regression, Regression Analysis, Standard Deviation, Variance, per capita

Click to edit the document details