This preview shows page 1. Sign up to view the full content.
Unformatted text preview: Economics 140A Fall 2011
Professor Startz
Answers to Homework 2 1.
a)
b)
c) √ . 2.
a) ∑
̅ ∑
̅ ̅ ̅ b)
∑ ̅ ∑ ̅ ̅ ∑ ̅∑ ̅ ̅̅̅ ∑ ̅̅̅ ∑
c) ∑ ̅ ̅ ∑ ̅∑
̅ ̅ ∑
̅ d)
∑ ̅ ̅ ∑ ∑ ∑
̅̅ ∑ ̅ ̅̅ ∑ ̅ ∑ ̅̅ ̅̅
̅̅ 3.
a) The sample mean is . The sample median is the value in the middle, 2.0. The sample variance is
4. Since the data is normal, so is the mean. Since the normal is a continuous distribution, the
probability of getting exactly a given value is effectively zero.
5.
a)
, since ̅ page 2
b) c) √ √ √ √ √
√
In general, any two variables that are linear functions of a third random variable are
perfectly correlated.
6. That was fun to read.
7.
a) Yes 8. Yes. This was really just practice in reading in an Excel spreadsheet.
9.
10. The output looks something like
120,000 Se r ies: WSAL_ VAL
Sa mple 1 133092
O b s e r v a t io n s 1 3 3 0 9 2 100,000 Mean
Median
Maximum
Minimum
Std. Dev.
Skewness
Kurtosis 60,000 40,000 20,000 29321.03
17000.00
1699999.
0.000000
47558.07
8.081066
142.1082 JarqueBera
Probability 80,000 1.09e+08
0.000000 0
1 250001 500001 750001 1000001 1250001 1500001 The statistics were Mean= 29,321 Median = 17,000, Maximum = 1,699,999, and Minimum =
0. The first three make sense as annual income; the zero minimum is probably for people
who aren’t working. page 3
11. For women the results were
60,000 Se r ie s: WSAL _ VAL
Sample 1 133092 IF F E=1
O b s e r v a t io n s 6 8 7 7 8 50,000 Mean
Median
Maximum
Minimum
Std. Dev.
Skewness
Kurtosis 30,000 20,000 10,000 2 1892.41
1 2000.00
1 099999.
0 .000000
3 4002.86
8 .437943
1 95.1424 JarqueBera
Probability 40,000 1 .07e+08
0 .000000 0
1 200001 400001 600001 800001 1000001 and for men they looked like
50,000 Se r ie s: WSAL _ VAL
Sample 1 133092 IF F E=0
O b s e r v a t io n s 6 4 3 1 4 40,000 Mean
Median
Maximum
Minimum
Std. Dev.
Skewness
Kurtosis 30,000 20,000 3 7265.26
2 5000.00
1 699999.
0 .000000
5 7636.73
7 .272244
1 07.8389 10,000 JarqueBera 3 0020541
Probability
0 .000000
0
1 250001 500001 750001 1000001 1250001 1500001 Not too surprisingly, the men get paid more. (One might want to remember that the data
includes people not working.) The average difference is around $15 ,000. 11. We have
̂ ̂ ̂
̂ ̂ (( )) page 4
12. The variance equals The standard deviation is √
. When I did it in EViews I got 0.0386, which was actually a little closer
than I expected.
13. I opened the variable FE and chose Descriptive statistics/simple hypothesis tests. The results
looked like this.
Hypothesis Testing for FE
Date: 09/26/11 Time: 16:34
Sample: 1 133092
Included observations: 133092
Test of Hypothesis: Mean = 0.500000
Sample Mean = 0.516770
Sample Std. Dev. = 0.499721
Method
tstatistic Value
12.24309 Probability
0.0000 The 95 percent critical value is 1.96, but the computed test statistic is so big, 11.28, that the
hypothesis would be rejected at any reason level of significance.
The critical value for a 99 percent confidence interval is 2.57, so the confidence interval is
.
√ ...
View Full
Document
 Fall '08
 Staff
 Economics

Click to edit the document details