Fall 2005 Homework Assignments.
(HW 1) Due Mon., Sept. 19, 2005:
Ch. 2, Problems 1, 5, 8, 12, 19, 20, 24.
In problem #5 of this Homework, you should assume that
N is very large, so that
the finite-population correction
factor 1-n/N is treated as 1.
(HW2) Due Wed., Oct. 5, 2005:
Ch. 2, #13; Ch. 3, #2, 3, 5, 10, 15, 17.
When you try to input the ASCII dataset
golfsrs.dat from the Lohr CD-rom
using
read.table in R, you will fail unless you first
delete the character "#" in line 89.
(HW3) Due Mon., Oct. 17, 2005:
Ch. 4, #3, 5, 6, 17, 21; Ch. 5, #3.
NOTES: In problem #3 of Chapter 4, you
must find the stratum population size
Nh by extrapolating
from stratum area using the given information about
"area per dredge". In #5, you are to do optimal allocations, and
nothing is said
about per-observation costs, so you should
assume they are equal across strata.
In #17, as long as the
roads and woods are roughly uniformly distributed over the
km^2 area, the stratum proportions in the systematically
sampled areas near
grid-line intersections should be roughly
the same as the stratum proportions in
the whole area.
(HW4) Due Mon., Oct. 31, 2005:
Ch. 5, #4, 9, 12, 15, 17, 22; Ch. 6, #4.
Remarks/hints about a few of the problems
(mostly written in response to
students' emailed questions) can be
found here .
(HW5) Due Wed., Nov. 23, 2005:
Ch. 6, #3, 5, 8, 12, 13, 16, 18.
(HW6) Due Mon., Dec. 12, 2005:
Ch. 3, #20; Ch. 7, #16; Ch. 8, #8; Ch. 9, #4, 9, 11.
Note that in Ch.3 #20(a), the "weighted least squares estimator"
is just
the ordinary least squares estimator (for slope,
when intercept is
constrained to be 0).