PADM 7
2
13
Final Project
Due Wednesday, December 7 by 11:50 PM
Note: this can be used as an artifact for
NASPAA Competency 3: Analyze, synthesize, think creatively, solve problems, and make decisions.
This assignment is designed to demonstrate your ability to:
1) Analyze scholarly research aimed at informing public policy
2) Correctly interpret and synthesize results of complex statistical analysis
3) Think creatively about how changes to research design could impact findings
4) Discuss how research findings inform decision making and problem-solving
Please read the following article:
Houston, D. J., & Richardson Jr, L. E. (2004). Drinking-and-driving in America: A test of behavioral assumptions underlying public policy.
Political research quarterly,
57(1), 53-64.
I.
Analyze the Framework
1) What is the main question that the article is attempting to answer? USE YOUR OWN WORDS (1 paragraph please).
The main question the article is seeking to answer is whether policies based on deterrence theory are effective against drunk-driving. It seeks to test the behavioral assumptions underlying the use of deterrence theory towards drinking and driving and seeks to establish the extent to which these public policies are significant.
2) How do social definitions of behavior such as deviant and “sinful” affect the choice of policy to address those behaviors? USE YOUR OWN WORDS (1 paragraph).
Social definitions of behavior using words such as deviant and sinful are likely to result in the enactment of policies that are punitive. Law enforcement agencies are more likely to implement sanctions aimed at discouraging such behavior. However, the punitive policies are less likely to be successful and hence alternative approaches should be considered when making the decisions on the policies to use.
II.
Interpretation of Statistical Analysis
The main variables of interest for the authors are two dummy variables:
·
Occasional drink-driver: =1 if an occasional drink-driver, 0 otherwise (base case is never drink-driver)
·
Frequent drink-driver: =1 if a frequent drink-driver, 0 otherwise (base case is never drink-driver)
Table 1: Please refer to Table 1 for the following questions:
Table 1: BAC Model
1) What is the dependent variable in this model? Please tell me what the coding is. i.e. =1 if……., 0 otherwise.
knows the legal BAC: =1 if knows the legal BAC levels; 0 otherwise
2) Simple interpretation of raw logistic coefficient Occasional drink driver. Also report statistical significance.
The raw logistic coefficient occasional drink driver is 0.434. The occasional drink drivers are more likely to know their legal BAC levels than non-drink drivers. The result is significant. p<0.01.
3) Interpretation of OR Occasional drink driver (percentages, please)
The odds ratio for occasional drink driver is 1.543. This means they are 1.543 times more likely to know their legal BAC levels than not knowing their legal BAC levels. This means that of occasion drink-drivers, 60.7% know their legal BAC levels while 39.3% do not know their legal BAC levels.
4) Simple interpretation of raw logistic regression coefficient frequent drink driver Also report statistical significance.
The raw logistic coefficient frequent drink driver is 0.476. The frequent drink drivers are more likely to know their legal BAC levels than non-drink drivers. The result is significan. p<0.05
5) Interpretation of OR for frequent drink-driver (percentages, please)
The odds ratio for frequent drink driver is 1.608. This means they are 1.608 times more likely to know their legal BAC levels than not knowing their legal BAC levels. This means that of frequent drink-drivers, 61.7% know their legal BAC levels while 38.3% do not know their legal BAC levels.
6) Simple interpretation of raw logistic regression coefficient on age. Also report statistical significance.
The raw logistic coefficient on age is -0.017. As age increases, the knowledge of BAC levels goes down. The results is significant. p<0.01
7) Interpretation of OR for age (percentages, please)
The OR for age is 0.983. This means when age is considered, that people are 0.983 times likely to know their legal BAC levels than not knowing their legal BAC levels. It means that when age is considered, 49.6% know their legal BAC levels while 50.4% do not know their legal BAC levels.
Table 1: Likely to be stopped Model
1) What is the dependent variable in this model? Please tell me what the coding is. i.e. =1 if……., 0 otherwise.
likely to be stopped:=1 if Likely to be stopped, 0 otherwise
2) Simple interpretation of raw logistic coefficient Occasional drink driver. Also report statistical significance.
The raw logistic coefficient occasional drink driver is -0.531. The occasional drink drivers are less likely to believe they will be stopped than non-drink drivers. The result is significant. p<0.01.
3) Interpretation of OR Occasional drink driver (percentages, please)
The odds ratio for occasional drink driver is 0.588. This means they are 0.588 times less likely to believe they will be stopped than not believing they are likely to be stopped. This means that of occasion drink-drivers, 37.0% believe they are likely to be stopped while 63.0% do not believe they are likely to be stopped.
4) Simple interpretation of raw logistic regression coefficient frequent drink driver Also report statistical significance.
The raw logistic coefficient frequent drink-driver is -0.518. The frequent drink drivers are less likely to believe they will be stopped than non-drink drivers. The result is significant since p<0.05
5) Interpretation of OR for frequent drink-driver (percentages, please)
The odds ratio for frequent drink driver is 0.596. This means they are 0.596 times less likely to believe they will be stopped than not believing they are likely to be stopped. This means that of frequent drink-drivers, 37.3% believe they are likely to be stopped while 62.7% do not believe they are likely to be stopped.
**Extra Credit!
On page 58, the authors write the following:
“Frequent drink-drivers and occasional drink-drivers are about half as likely as non-drink-drivers to think it is “almost certain” or “very likely” they would be stopped.”
Why is this interpretation
wrong? (Hint: think about risk ratios versus odds ratios). (A few sentences)
The interpretation is wrong because the authors seemingly divided the odd ratios of frequent drink-drivers and occasional drink-drivers by the odd ratio for the non-drink drivers to arrive at the figure of about half. A more accurate approach would have been to calculate the ratios of each group and make a comparison.
Table 1: Almost certain to receive punishment Model
1) What is the dependent variable in this model? Please tell me what the coding is. i.e. =1 if……., 0 otherwise.
Almost certain to receive punishment: =1 if Almost certain to receive punishment, 0 otherwise
2) Simple interpretation of raw logistic coefficient Occasional drink driver. Also report statistical significance.
The raw logistic coefficient occasional drink driver is 0.247. The occasional drink drivers are less likely to believe they will be stopped than both frequent drinkers and non-drink drivers. The result is significant. p<0.05.
3) Interpretation of OR Occasional drink driver (percentages, please)
The odds ratio for occasional drink driver is 1.280. This means they are 1.280 times more likely to be almost certain to receive punishment than not almost being certain to receive punishment. This means that of occasion drink-drivers, 56.1% are almost certain to receive punishment and 43.9% are not almost being certain to receive punishment.
4) Simple interpretation of raw logistic regression coefficient frequent drink driver Also report statistical significance.
The raw logistic coefficient frequent drink driver is 0.278. The frequent drink drivers are less likely to believe they will be stopped than non-drink-drivers but more than occasional drinkers. The result is significant. p<0.10.
5) Interpretation of OR for frequent drink-driver (percentages, please)
The odds ratio for frequent drink driver is 1.320. This means they are 1.320 times more likely to be almost certain to receive punishment than not almost being certain to receive punishment. This means that of frequent drink-drivers, 56.9% are almost certain to receive punishment and 43.1% are not almost being certain to receive punishment.
Table 1: Punishment very severe Model
1) What is the dependent variable in this model? Please tell me what the coding is. i.e. =1 if……., 0 otherwise.
Punishment very severe
: =1 if
punishment very severe, 0 otherwise
2) Simple interpretation of raw logistic coefficient Occasional drink driver. Also report statistical significance.
The raw logistic coefficient occasional drink driver is 0.093. The occasional drink drivers are less likely to believe they will be stopped than both frequent drinkers and non-drink drivers. The result is NOT significant
3) Interpretation of OR Occasional drink driver (percentages, please)
The odds ratio for occasional drink driver is 1.098. This means they are 1.098 times more likely to believe the punishment will be severe than not to believe the punishment will be severe. This means that of occasion drink-drivers, 52.3% believe the punishment will be severe and 42.7% do not believe the punishment will be severe.
4) Simple interpretation of raw logistic regression coefficient frequent drink driver Also report statistical significance.
The raw logistic coefficient frequent drink driver is 0.284. The frequent drink drivers are less likely to believe they will be stopped than non-drinkers but more likely than occasional drink drivers. The result is significant since p<0.10
5) Interpretation of OR for frequent drink-driver (percentages, please)
The odds ratio for occasional drink driver is 1.329. This means they are 1.329 times more likely to believe the punishment will be severe than not to believe the punishment will be severe. This means that of occasion drink-drivers, 57.1% believe the punishment will be severe and 42.9% do not believe the punishment will be severe.
Table 1 Critical appraisal: Given the results in Table 1, what conclusions do you reach about drink-driving attitudes and behaviors? Think about what each regression is telling you about drink-drivers USE YOUR OWN WORDS (1 paragraph, please)
They use of deterrent policies have no major effect on drink-driving attitudes and behaviors. As can be seen in the table, the 4 logistic regression models failed to show that the deterrent policies had any major effect on drunk-driving.
Table 2: Please refer to Table 2 for the following questions:
Table 2: Police stop more likely than an accident Model
1) What is the dependent variable in this model? Please tell me what the coding is. i.e. =1 if……., 0 otherwise.
Police stop more likely than an accident: =1 if Police stop more likely than an accident Model, 0 otherwise
2) Simple interpretation of raw logistic coefficient Occasional drink driver. Also report statistical significance.
The raw logistic coefficient occasional drink driver is 0.306. The occasional drink drivers are less likely to believe that police stop are more likely than an accident than both the non-drinking-drivers and the frequent drivers. The result is significant. p<0.01
3) Interpretation of OR Occasional drink driver (percentages, please)
The odds ratio for occasional drink driver is 1.358. This means they are 1.358 times more likely to believe that police stop are more likely than an accident than they are not likely to believe that police stop are not more likely than an accident. This means that of occasion drink-drivers, 57.6% believe that police stop are more likely than an accident and 42.4% do not believe that that police stop are more likely than an accident.
4) Simple interpretation of raw logistic regression coefficient frequent drink driver. Also report statistical significance.
The raw logistic frequent frequent drink driver is 0.403. The frequent drink drivers are more likely to believe that police stop are more likely than an accident than both the non-drinking-drivers and the occasional drink drivers. The result is significant. p<0.05
5) Interpretation of OR for frequent drink-driver (percentages, please)
The odds ratio for frequent drink driver is 1.497. This means they are 1.497 times more likely to believe that police stop are more likely than an accident than they are not likely to believe that police stop are not more likely than an accident. This means that of frequent drink-drivers, 60.0% believe that police stop are more likely than an accident and 40.0 % do not believe that that police stop are more likely than an accident
**Extra Credit!
On page 60, the authors write:
“The odds ratios indicate that frequent drink-drivers are one and a half times more likely to offer this response than non-drink-drivers.”
Why is this interpretation
wrong? (Hint: think about risk ratios versus odds ratios). (A few sentences)
The interpretation is wrong because the authors seemingly divided the odd ratios of frequent drink-drivers by the odd ratio for the non-drink drivers to arrive at the figure of about one and half times. A more accurate approach would have been to calculate the ratios of each group and make a comparison.
Table 2: Major threat to personal safety Model
1) What is the dependent variable in this model? Please tell me what the coding is. i.e. =1 if……., 0 otherwise.
Major threat to personal safety: =1 if
Major threat to personal safety: =1 if Major threat to personal safety, 0 otherwise
2) Simple interpretation of raw logistic coefficient Occasional drink driver. Also report statistical significance.
The raw logistic coefficient occasional drink driver is -0.585. The occasional drink drivers are more likely than non-drinking drivers but less likely than frequent drink drivers to have major threat to personal safety model
3) Interpretation of OR Occasional drink driver (percentages, please)
The odds ratio for occasional drink driver is 0.557. This means they are 0.557 times more likely to have a major threat to personal safety than not to have a major threat to personal safety. This means that of occasion drink-drivers, 35.8% have a major threat to personal safety than not to have a major threat to personal safety while 64.2% do not have a major threat to personal safety.
4) Simple interpretation of raw logistic regression coefficient frequent drink driver Also report statistical significance.
The raw logistic coefficient frequent drink driver is -0.772. The frequent drink drivers are more likely than both non-drinking drivers and occasional drink-drivers to have major threat to personal safety
5) Interpretation of OR for frequent drink-driver (percentages, please)
The odds ratio for frequent drink driver is 0.462. This means they are 0.462 times more likely to have a major threat to personal safety than not to have a major threat to personal safety. This means that of frequent drink-drivers, 31.6% have a major threat to personal safety than not to have a major threat to personal safety while 68.4% do not have a major threat to personal safety.
**Extra Credit!
On page 60, the authors write,
“Both groups of drink-drivers are about half as likely as non drink-drivers to see this behavior as a “major” threat to personal safety.”
Why is this interpretation
wrong? (Hint: think about risk ratios versus odds ratios). (A few sentences)
The interpretation is wrong because the authors seemingly divided the odd ratios of both groups of drink-drivers by the odd ratio for the non-drink drivers to arrive at the figure of about half. A more accurate approach would have been to calculate the ratios of each group and make a comparison.
Table 2: Serious threat to safety Model
1) What is the dependent variable in this model? Please tell me what the coding is. i.e. =1 if……., 0 otherwise.
Serious threat to safety: =1 if serious threat to safety, 0 otherwise
2) Simple interpretation of raw logistic coefficient Occasional drink driver. Also report statistical significance.
The raw logistic coefficient occasional drink driver is -0.412. The occasional drink drivers are more likely than non-drinking drivers but less likely than frequent drink drivers to have serious threat to personal safety model
3) Interpretation of OR Occasional drink driver (percentages, please)
The odds ratio for occasional drink driver is 0.663. This means they are 0.663 times more likely to have a serious threat to personal safety than not to have a serious threat to personal safety. This means that of occasion drink-drivers, 39.9% have a serious threat to personal safety while 60.1% do not have a serious threat to personal safety.
4) Simple interpretation of raw logistic regression coefficient frequent drink driver Also report statistical significance.
The raw logistic coefficient frequent drink driver is -0.828. The frequent drink drivers are more likely than both non-drinking drivers and occasional drink-drivers to have serious threat to personal safety model
5) Interpretation of OR for frequent drink-driver (percentages, please)
The odds ratio for frequent drink driver is 0.437. This means they are 0.437 times more likely to have a serious threat to personal safety than not to have a serious threat to personal safety. This means that of frequent drink-drivers, 30.4% have a serious threat to personal safety while 69.6% do not have a serious threat to personal safety.
Table 2 Critical appraisal: Given the results in Table 2, what conclusions do you reach about drink-driving attitudes and behaviors? Think about what each regression is telling you. USE YOUR OWN WORDS (1 paragraph, please)
Drink-driving attitudes and behaviors cause both major and serious threats to personal safety. From the coefficients and odd ratios, this is clearly evident. On safety, frequent drink drivers pose the most serious risk followed by occasional drink-drivers and last is non drinking-drivers.
Table 3: Please summarize the policy preferences of occasional and frequent drink-drivers in Table 3. I want you to look at each model and then summarize the patterns of policy preference for drink-drivers versus non drink-drivers for each policy preference. You can refer to the coefficients or not in your summary-it’s up to you. USE YOUR OWN WORDS. (1-2 paragraphs).
Drink drivers are more likely than non-drink-drivers to think of the current laws are very effective. Of the three groups, frequent drink drivers are the least likely to think of the need for stringent policies while the non-drive-drinkers are the most likely to think that stringent rules are needed. In most cases, frequent drink-drivers are less supportive of stringent policies while non drink-drivers are more support the stringent policies than the drink-driving group.
III.
Overall Paper Assessment
1) In your own words, what are the policy implications of this article? That is, based on the findings, what would you recommend in terms of policy? USE YOUR OWN WORDS (1-2 paragraphs please).
Based on the article, deterrent policies have limited effects in eliminating drink-driving behavior and other sinful behavior. Alternative strategies should be put in place and taking into consideration the diversities of various groups. Alternative strategies may more likely be more successful than deterrent policies though more research need to be done on their effectiveness in reducing the frequency of drunk-driving.
2) Please identify one or more conclusions in the study with which you DISAGREE. Please explain why you disagree with the conclusion and your alternative conclusion. (1-2 paragraphs).
I disagree with the conclusion that deterrent policies are less effective. More deterrent policies like heavy fines and jail terms would most likely cause more drink-drivers to be more responsible to avoid the negative repercussions of their drunk behaviors. My alternative conclusion is that other alternative approaches should not replace the deterrent policies, but should complement them to have an overall reduction in drunk-driving.
2
Instructions for Critical thinking paper
NASPAA Competency: To Analyze, synthesize, think critically, solve problems, and
make decisions (artifact: PADM 7601, 7602, 7612, 7213)
This competency means that students should be able to demonstrate the ability to do the
following:
1. Identify a problem or research questions
2. Collect and analyze data or information related to that problem
3. Draw meaningful conclusions from that analysis that informs decision-making or
solutions
Artifact: {Link to paper or project which demonstrates knowledge of this competency. You
will upload your paper and link it on the “Artifact” tab for this page.}
1)
How does this artifact demonstrate your ability to identify a problem or research
questions?
(100-200 words)
{Text here}
2)
How does this artifact demonstrate your ability to collect and analyze data or
information? (100-200 words)
{Text here}
3)
How does this artifact demonstrate your ability to draw meaningful conclusions
from that analysis that informs decision making or solutions? (100-200 words)
{Text here}
4)
Discuss three ways in which your coursework in the MPA program has enhanced
your ability to analyze, synthesize, think critically, solve problems, and make decisions
(100-200 words)
Order an Essay Now & Get These Features For Free:
Turnitin Report
Formatting
Title Page
Citation
Outline
