Basic HTML Version
Q
uantitative
R
easoning &
P
roblem
S
olving
320
© 2014 Pacific Crest
M
ethodology
P
erforming
D
ata
A
nalysis
Scenario:
You are a newspaper reporter and your editor has asked you for a summary of
homicides for the last 20 years to be included with a story in tomorrow’s paper. She
specifically asked for neutral reporting of the impact guns have on homicides and
a comparison of homicides by guns with other homicides. Any trends you identify
over the last 20 or so years (since 1977 at least) would be welcome too. You find
the following data on Homicides by Weapon Type from the FBI Supplementary
Homicide Reports, 1976 – 2005 (link available at the companion website).
Handgun
Other
Gun
Knife
Blunt
Object
Other
Weapon
1977
8,563 3,391 3,648 900 2,618
1978
8,879 3,569 3,685 937 2,490
1979
9,858 3,732 4,121 1,039 2,710
1980
10,552 3,834 4,439 1,153 3,061
1981
10,324 3,740 4,364 1,166 2,927
1982
9,137 3,501 4,383 1,032 2,957
1983
8,472 2,794 4,214 1,098 2,731
1984
8,183 2,835 3,956 1,090 2,626
1985
8,165 2,973 3,996 1,051 2,794
1986
9,054 3,126 4,235 1,176 3,018
1987
8,781 3,094 4,076 1,169 2,980
1988
9,375 3,162 3,978 1,296 2,869
1989
10,225 3,197 3,923 1,279 2,877
1990
11,677 3,395 4,077 1,254 3,037
1991
13,101 3,277 3,909 1,252 3,161
Handgun
Other
Gun
Knife
Blunt
Object
Other
Weapon
1991
13,101 3,277 3,909 1,252 3,161
1992
13,158 3,043 3,447 1,088 3,024
1993
13,981 3,094 3,140 1,082 3,233
1994
13,496 2,840 2,960 963 3,071
1995
12,050 2,679 2,731 981 3,169
1996
10,731 2,533 2,691 917 2,777
1997
9,705 2,631 2,363 833 2,678
1998
8,844 2,168 2,257 896 2,805
1999
7,943 2,174 2,042 902 2,461
2000
7,985 2,218 2,099 727 2,556
2001
7,900 2,239 2,090 776 3,032
2002
8,286 2,538 2,018 773 2,588
2003
8,830 2,223 2,085 745 2,645
2004
8,304 2,357 2,133 759 2,595
2005
8,478 2,868 2,147 671 2,528
Step
Explanation
1.
Clarify the data
What does each variable and row represent? Are there issues of quality
such as errors or missing data? What are the values being measured?
What are the relevant units? Are there any outliers that were removed? Is
bias a problem?
WATCH
IT WORK!
Death by
handgun
, any
other gun
,
knife
,
blunt object
,
other weapon
. Data is
annual national deaths collected by the FBI; no missing data; no assumed errors.
2.
Clarify the context
Who is asking for the analysis, what is the agenda, and what are the types
of results that they want or expect to see? Identify the objectives (3 to 5)
and always consider possible reasons for bias in the reporting of this data.
Your editor wants a quick summary of homicides for the last 20 years to educate
readers in tomorrow’s paper. You have just 2 hours to do this analysis. You’ve been
asked for neutral reporting of the impact guns have on homicides. You must give a
clear presentation, compare homicides by guns with other homicides, and note any
trends over the last 20 years.