Qualifiers

[BG45]

[center][b]Policy Note Cards: Sex Offenders[/b][/center]

Bandy (2011) – Sex offenders are concentrated in disadvantaged neighborhoods. Disclosing offender info has no impact on people’s self protection except for kids.

Sample (2011) – Little evidence SORN reduces re-offense. Symbolic laws are rarely repealed.

Terry (2011) – Adam Walsh Act. Juveniles can now be sex offenders for life. Eliminates mitigating circumstances. Failure to register is now a federal crime.

Zgoba (2011) – Every offender still has access to public transit despite residency restrictions. 700,000 registered sex offenders in the US, 100,000 are missing. Residency restrictions are ineffective.

Tewksbury (2011) – Sex offenders forced by SORN into poor or rural communities. Not all victims are kids, which is what residency restrictions address.

Socia (2011) – Residency restrictions in upstate NY don’t help rehabilitation or public safety.

Barnes (2011) – Wants moratorium on SORN, residency restrictions don’t work in Minnesota study.

Levenson (2001) – Sex offender recidivism is lower than the average criminal. Punishment for crimes they might commit, instead of what they do commit.

Harris et al. (2010) – Public attitude to sex offenders is driven by media myths.

Levenson et al. (2010) – Whether a sex offender fails to register has no effect on recidivism.

Burchfield (2011) – No assumption in SORN residency restrictions is backed by empirical evidence. Anecdotes of sex offender colonies due to effects of SORN.

Hakim (2011, personal communication) – Sex offender colonies under bridges in some California cities due to residency restrictions.

Leon (2011) – Plea deals can place high risk offenders in low risk categories.

[BG45]

[center][b]Policy Note Cards: Immigration[/b][/center]

Immigration and Naturalization Act (1952) – first modern immigration reform.

Immigration Reform and Responsibility Act (1986) – Increase Border Patrol funding, 1 time amnesty if you were in the US before 1982.

Illegal Immigration Reform and Responsibility Act (1996) – More BP funding, easier to arrest and deport non citizens.

Secure Border Initiative (2006) – More BP funding, more wall funding.

Arizona law SB1070 (2010) – Most of US supports. Ruled unconstitutional 3 days after enforcement began.

Stowell et al. (2009) – Chicago schools with a large immigrant population had less crime. Time series and metropolitan areas between 1994-2004 showed an increase in immigrant populations and a decrease in crime.

Bucerius (2010) – Associating immigrants with crime and viewing them with distrust may contribute to weak social interaction. High crime by the 2[sup]nd[/sup] generation may be due to acculturation or perceiving prejudice and inequality different than their parent’s group. Suggests policies that reduce social exclusion, focus on school experiences.

Martinez (2011) – Latinos are ignored in research and latino violence decreased, despite immigration being up.

Martinez (2010) - Case study of immigrants and crime, are border areas more violent due to undocumented workers? Homicide levels were lower in border communities, inverse relationship found between immigrant concentration and lethal violence. Growing literature shows reform needed to stop targeting immigrants. Social structure factors most important (disadvantaged neighborhoods, residential instability). Also, less trust of cops leads to them being less willing to report a crime.

Sutherland (1924) – Challenged the perception that immigrants commit more crime. Argue that crime was an effect of acculturation into US society. 2[sup]nd[/sup] generation tends to be more criminal.

Bahrampour (2011) – ICE Secure Communities Program. Arrests run against immigration database.

Merton (1938) – Viewed immigrants as conformists.

Decker (2010) – Hard construct crime for illegal immigrants, don’t know denomination. Arizona approach could strain cop-immigrant relations more, more distrust means that there is less reporting. Police chiefs often disagree with public and political action. Should not use criminal law to address a problem beyond its scope.

Hickman & Suttorp (2008) – Earlier period immigrants more favorable, intensified by 9/11. More panic = less tolerance of immigrants. In 2008, sense of urgency led to many actions on Mexican border, 700 mile fence, 100 new BP agents, 6k national guard. Fear of illegals increasing crime, no evidence that illegals are a unique threat.

Hagan et al. (2008) – Acculturation. 2[sup]nd[/sup] generation more criminal. 2006, Hazelton PA, wave of city council ordinances declared illegals lead to higher crime. 75% of Americans agree. Law would fine businesses that hire illegals and penalize landlords who rent to them, struck down by US District Court.

Hagan & Phillips (2008) – US continued hard on immigrants, soft on businesses policies. Flow of policy initiatives dramatically increase deportation, over 200,000 in 2008 alone under Obama administration. Mass deportation has caused big policy issues: Immigrants are often mistreated, no way to report abuse; many removed are long term and leave family behind; stereotype of immigrant criminal common. US generally ignores the Canadian border. Policy implications: Immigration policy must include guest worker program, enforced employer scrutiny, legalization path of immigrants who have established work and family ties.

[BG45]

[b]Policy Note Card Addendum: Human Trafficking[/b]

Passion Conference 2012, a Christian youth gathering, raised over $3 million in immediate donations to help NGOs fight human trafficking.

[brianthephysicist]

Wow! Man, this is an impressive amount of information! Go show them how boss you are!

[BG45]

Thanks Brian! I hope to type out the two Theory sections up tonight that I forgot in PA when I was home. I've had a professor who passed all three the first time (and who sat on one committee that failed me last time) reviewing my practice answers along with one woman who just finished her PhD in our program and one of my cohort members who passed all three the first time.

I'm hoping and praying that God's got my back, and was just humbling me the first time...

[BG45]

[center][b]Theory Note Cards: Social Control Theories [/b][/center]

Sykes & Matza (1957)/Matza (1964) – Propose individuals free themselves from constraint based on techniques of neutralization. Matza stated most of the time delinquents are not engaged in delinquency and most age out. Delinquents are not committed to delinquent beliefs. Most delinquents drift in and out of delinquency based on fluctuations in control.

Kemph (1993) – Review of 71 studies on control, most supported. However, variables were operationalized differently. When the variables were examined individually, the results were mixed and inconsistent. This lead to the conclusion that these were essentially separate studies, which failed to build on experience. Little revealed about the viability of social control.

Hirschi (1969) – Social bond. Delinquent act from when bond to society weakens. There are 4 areas to bond to society. Attachment, the level of parental supervision. Involvement. Commitment, school and occupational aspirations. Belief, in the rules of society. When one or more bonds weakened, crime is more likely. Tested with self report of 4,000 high schoolers, found support.

Krohn & Massey (1980) – tested social bond. Examined how well variables explained pot and booze use and minor delinquency. Found the variables to explain minor forms better than serious forms.

Jensen & Brownfield (1983) – Evidence against attachment = less delinquent. Even if attached, can be delinquent.

Kubrin, Stucky, and Krohn (2009) – Cross sectional support, but longitudinal support is lacking.

Agnew (1985) – longitudinal with Youth in Transition survey data. Looked at felony and minor assault and theft. Hypothesized people with strong bonds to non-dangerous ideals and committed to beliefs and goals would commit less crime. Found moderate relationship. Association with delinquent peers is the best predictor.

Laub & Sampson (1988) – Reanalysis of the Glueck data. Social control via harsh discipline, weak parental supervision, parental rejection of child, and weak emotional attachment of a boy to parents results in harsh delinquency. Combined with structured factors using OLS.

Agnew (1991) – Longitudinal, 2 waves, n= 49. While delinquent peers and beliefs increased probably of delinquency, parental attachment only indirectly correlated. Used National Youth Survey data.

Durkheim (1895) – Asserted every society would have crime. Crime serves as a boundary maintenance function where human behavior is controlled by social reaction. Link between anomie and social control, as norms break down, so does control.

Reiss (1951) – Delinquency occurs when 1 of the following takes place: Lack of proper internal controls developed in youth, breakdown of those controls, absence of societal rules or norms. These elements are now in most control theories.

Barbara et al. (2009) – Most studies of social control conclude it is supported by the data.
Toby (1957) – Stakes in conformity show much to lose. Less to lose = more crime.

Assumptions of Social Control Theory – Both social and self control. Assume people are naturally inclined to deviate but are constricted due to self control or bonds to society.

[BG45]

[center][b]Theory Note Cards: Social Learning Theories [/b][/center]

Sutherland (1949) –White collar crime – why do some associations lead to deviance and others do not? Used frequency, priorities, and case histories.

Burgess & Akers (1966)/Akers (1973,1985,1998) – 4 components. A) Definitions (general and specific) B) Differential Association C) Differential reinforcement, usually measure how others will view behavior, but some expect consequence and reward. D) Imitation.

Akers et al. (1979) – Study examined the 4 principles of social learning. 3,000 Midwest students, grades 7-12. DV alcohol and drug use and abuse. Combined model of the 4 variables accounted for 55% of variance in alcohol and 68% in week. In terms of abuse, 32% and 39%. All variables had positive impact for increased delinquency.

Matsueda (1982) – Tested definitions from differential association. Continuum from favorable ________unfavorable. More definitions = more favorable = higher delinquency. More support for DA than social control. Used structural equation modeling.

Krohn et al. (1985) – Most longitudinal studies of social learning show support for the theory.

Matsueda (1989) –responded to early criticisms of his work, stated theory can be tested. Past studies show support for the theory.

Agnew (1991) – Learning vs. control. Found more support for learning.

Warr & Stafford (1991) – test to see if peer attitudes or peer behavior is more important. Used NYS data. Both attitudes and behavior were significant in determining one’s own behavior. Friend behavior was 5 times greater impact than friend’s attitude, 2.5 times more so than own attitude. In a log. Analysis, the impact of delinquent peers stayed strong. Attitude is less important than behaviors.

Akers and Lee (1999) – Imitation was too weak a correlate, they finally dropped it from the model.

Costellle and Vowel (1999) – Stated that Matsueda measured control wrong.

Pratt and Cullen (2000) – Meta analysis. Found support for impact of 2 social learning variables (differential association and definitions) on offending, but social control was stronger.

Haynie (2002) – Looked at peer networks and matched child with friends, asking to list closest friends, match up corresponding. 7-12 graders, NH Longitudinal Adolescent Health Survey. Number of delinquent friends is better indicator than overall acts from network.
Chappell & Piquero (2004) – ask about specific consequences of cop misconduct, probably better measure than asking what you think a reaction to a hypothetical would be.

[BG45]

Tables won't show up But practice question I just finished for afternoon Methods, it's the one i failed as an actual Qualifier in August. (Also PM likes to renumber my stuff...)

[b]Question:[/b]

[center][b][u]RESEARCH METHODS[/u][/b][b][u] Qualifying Exam[/u][/b][/center]
[center][b][u]Afternoon Question –August 2011[/u][/b][/center]
[b]Instructions: [/b]You have two (2) hours to answer the afternoon questions. Answer each part of the question and label your answers (1, 2, 3, …).
[b]Question #3: Quantitative Analysis (34% of grade)[/b]
Researchers often contend that when the economy is stagnant or poor crime rates should increase. A measure of the economy is the unemployment rate. A researcher has collected data on the violent crime rate in each state along with the unemployment rate.
Interpret the following tables:[list=1]
[*]State a [b]two tailed[/b] research hypothesis for the dependent and independent variables.
[*]Compute and interpret the coefficient of determination.
[*]Interpret the unstandardized and standardized coefficients. Demonstrate you know what they mean, and discuss the uses of each measure. Which do you think would be the better measure to use for these data? Justify your choice.
[/list][list=1]
[*]Construct and interpret the 95% confidence interval for the unstandardized coefficient. Demonstrate that you understand the process of statistical inference.
[/list][list=1]
[*]Explain whether or not the findings support your research hypothesis.
[/list][list=1]
[*]You decide to recode unemployment rate into unemployment rates as “above or the same as national unemployment average” or “below the national unemployment average” Interpret the following table. Once you have completed the recode, what type of analysis would you conduct to determine the relationship between this recoded variable and violent crime rate?
[/list]

[b]Answer: [/b]
[list=1]
[*]Hypothesis: There is a statistically significant relationship between unemployment and violent crime rates.
[/list][list=1]
[*]The coefficient of determination is known as R-squared. There are two ways of computing this; the first is to take the Sum of Squares regression (220474.194) and divide it by the Total Sum of Squares (1232343.446). This gives us 0.1789 or 17.89%. Another method of determining the R-square, and will now be used to double check, is to square the Beta. In an OLS regression, the Beta is equal to R, therefore Beta-squared will give us R-squared. Beta is .423, when we square this we get .1789; again we see that r-squared is .1789, or 17.89%.

Therefore, for this sample, by taking unemployment rate into account when predicting the violent crime rate, we are able to reduce prediction error by 17.89%, compared to the error made when the unemployment rate was not taken into account.
[/list][list=1]
[*]Unstandardized and Standardized Coefficients.
[/list]

The unstandardarized coefficient (or slope) is b, and is 34.459. For this sample, on average, a one unit increase in the unemployment rate is associated with an increase of 34.459 units in the violent crime rate.

The standardized coefficient, beta, is .423. For this sample, on average, a one standard deviation unit increase in the unemployment rate is associated with a .423 standard deviation unit increase in the violent crime rate. In a study with multiple independent variables, those variables often have different units of measurement and need to be standardized in order to properly compare them. Beta converts the slopes of different measurements in those studies into a standardized unit of measure, standard deviation units, so that the relative strength might be compared.

For this study, the unstandardized coefficient, or slope, is the preferable measure, as we have no other independent variables. Therefore, a common unit of measurement is unnecessary.[list=1]
[*]The 95% confidence interval for the unstandardized coefficient.
[/list]

In order to properly estimate a confidence interval for the unstandardized coefficient (or population slope) we must meet a number of assumptions that are used in OLS regression. These include, but are not limited to: independent random sampling, interval level data, equality of variance/homoscedasticity, and that the variables are normally distributed across the population. If these assumptions are met, we are able to utilize the Central Limits Theorem in order to estimate the population slope as part of our confidence interval.

If we were to go to a population where the assumptions were met and took every possible random sample size of 50, calculated every possible slope and we plotted all of those possible slopes, we would have a distribution of sample slopes. The mean of this sampling distribution of slopes would be equal to the population mean, B. This distribution is theoretical however, because it is almost impossible to conduct the operations necessary to actually create it. As an empirical distribution, we know that it has known properties. One of those properties is that we can use the t-distribution for a given alpha level and degrees of freedom to find the points between which 95% of the cases lie. For this case, the alpha level is .05 and our df = n-k= 48. Since the value for 48 is not given on the t-table, we’ll use the value for df = 50 and the difference between the two values is very small. The t-table value for the alpha and the degrees of freedom is 2.008. The empirical rule therefore states that 95% of the slopes in our sampling distribution of slopes fall within 2.008 standard deviations of the mean (or true population slope, if one prefers).

If we reverse the empirical rule, we can reason that if we put our confidence interval of plus or minus 2.008 standard deviations around each of the slopes in our sampling distribution of slopes, then 95% of our confidence intervals will capture the true population slope. The 5% that do not are a Type One Error, where we claim to have captured the true slope when we actually did not.

We know, therefore, that we have one sample slope from somewhere in our sampling distribution of sample slopes. We can use the information that has been laid out to place a confidence interval around it and we have a 95% chance that we will capture the true population slope, while having a 5% chance of committing a Type One Error.

Our confidence interval around the unstandardized coefficient is expressed as:
b plus or minus t(sb)
34.459 plus or minus 2.008(10.655)
34.459 plus or minus 21.395

The Confidence Interval around the unstandardized coefficient/slope is 13.064______55.854

If our lower limit of 13.064 is true, then, for this sample, on average, a one unit increase in the unemployment rate is associated with a 13.064 unit increase in the violent crime rate. If the upper limit of 55.854 is true, then for this sample, on average, a one unit increase in the unemployment rate is associated with a 55.854 unit increase in the violent crime rate.

Both of these rates show an increase in the violent crime rate and a positive relationship between the unemployment rate and the violent crime rate. However this is an extremely wide confidence interval, therefore it is of questionable use for anything outside of stating that there is a positive relationship. We must also admit that there is a 5% chance of having committed a Type One Error by wrongly stating that we’ve captured the true population slope in this interval, when we may have not.[list=1]
[*]The hypothesis given at the beginning was: There is a statistically significant relationship between unemployment and violent crime rates.
[/list]

That is a research hypothesis, but we interpret our results in the form of a null hypothesis. Our null hypothesis would be: There is not a statistically significant relationship between the rate of unemployment and violent crime rates. If the null hypothesis holds that there is no relationship between variables, and thus the population slope is zero, then we see that the confidence interval above has not contained zero and therefore we are 95% certain that the population slope is a non-zero slope. Therefore, we will reject our null hypothesis of no relationship between the variables and we will acknowledge that there is a 5% chance we are committing a type one error.[list=1]
[*]Other ways of analyzing this data. The independent variable in the final table provided shows us that our independent variable has been recorded into an ordinal level. This is somewhat fortunate, because we could do a number of things with the IV as an ordinal variable with interval/ratio for theDV.
[/list]

We would most likely be interested in comparing the relationship between the new variable and the violent crime rate, so we would conduct an independent samples T-test to compare our average crime rate between the two groups.Normally distributed variables are one of the assumptions of this test and the table shows that the distribution of the IV is tolerably distributed for this (24 are below the national average, 26 are above) and that the two groups are equal in size.A t-test would show whether or not there is a statistically significant deifference in the mean crime rate of the two groups (Below and Above national average).

[BG45]

[b]Question:[/b]

[b]January 2009: Theory Qualifier[/b]


[b][u]Morning Question[/u][/b][b]:[/b]

For several decades, Robinson Crusoe was a castaway on an "uncharted" island. Respond to the following:[list]
[*]Could Robinson Crusoe commit a crime while alone on the island?
[/list][list]
[*]Present your understanding of the conceptualization of crime, the origin of law, and the benefit of criminological theory in predicting criminal behaviors.
[/list][list]
[*]Discuss how this understanding can be useful in identifying, understanding, and (if necessary) preventing victimless crime.
[/list][list]
[*]Be sure to support your response with appropriate references to relevant, academic theoretical discussions, essays, articles, and/or books.
[/list]



[b]Answer:[/b]

One of the most fiercely debated topics in the news media today and the court of public opinion is whether or not there is such a thing as a victimless crime (Surette, 2007). Criminologists are rarely consulted about this, and it is of therefore little surprise that criminologists’ opinions do not make the mainstream. In the course of this response, the idea that Robinson Crusoe was capable of committing a crime on an island alone with be explored, expounded upon, and discussed. This will include a discussion of the creation of laws and contract with society (Hobbes), the ancient origins of law (Kurlychek, 2004), the idea of crime as a universal constant in any society (Durkheim, 1895), and the benefits of criminological theory in predicting basic human behaviors via deterrence and social control theories.

In the scenario that has been put forward, a castaway, one Robinson Crusoe has been stranded on a deserted island. He will be there for a number of decades utterly alone, and the question at hand is whether or not he can commit a crime. In order to understand the nature of possibly criminality, we must first look at the views of crime in a historical context.

The world’s first written code of law was the Hammurabi Code, written approximately 4,000 years ago by the Babylonian ruler of the same name. The Hammurabi Code was a vast undertaking, especially as the world’s first known codified set of laws. It dealt with any number of issues, from ordering the death of any architect whose work collapsed and killed a person to the treatment of juveniles differently than their adult counterparts in terms of punishment. However, one thing it did not deal with, was the thought of “victimless crimes” such as Robinson Crusoe would be involved in (Kurlychek, 2004).

After this period, most law was interpreted in the form of the Scriptures and of sin. There was not a thought that law and society were related, but rather all law was related to the natural law of God and man as expressed by different faiths (Durkheim, 1893). The most common two forms of this are the Sharia Law of the Islamic faith and that of the Natural (or Eternal) Law of Christianity (Aquinas, 1274). Thomas Hobbes would establish the next major thought on the nature of crime that did not involve a deity exclusively with his seminal work [i]Leviathan[/i] (1651). Hobbes, would declare, that the nature of crime was a breaking of an unwritten and unspoken social contract with the state. The state provided services, protection, a home, and a culture for an individual, it was up to the individual to maintain this contract by being a good citizen, therefore not a criminal.

In the situation of Robinson Crusoe, the island acts as the state. However, the state government that is needed for the social contract to be made with is, in effect, Crusoe himself. He is the only person on the island, therefore, he is the island government by default. Therefore, one is lead to ask, under the theory of social contract obligations, if Crusoe is really capable of committing a crime? After all, as the state, he is capable of nullifying the criminality of his own actions via law formulation.

It is therefore that the theories of Emile Durkheim must be brought into play on the island. Durkheim would be another key voice in the Enlightenment period, and would be the first to truly show that crime was not necessarily a sin against God, but a transgression against the society within which one existed.

Durkheim (1895) stated that there is always going to be crime in a society, no matter how small. Furthermore, crime was not an unnatural occurrence, but rather it acted as a boundary feature to the society within which it existed.

One could contend then, that Crusoe had entered a state of anomie, or normlessness. He was ripped from the comfortable confines of his normal existence and faced an immense social and personal upheaval that he could never have predicted would occur. This threw him not only on to an island where he was a newcomer, but an island where he was alone with his own thoughts and turmoil. The normal laws of the society he had belonged to were no longer applicable to him and therefore he no longer held a contract with his former state; but even if he did find himself to be the state, that does not mean he was incapable of committing a crime in his normlessness.

Crusoe was a one man society, but as a one man society, he was still capable of creating the boundaries for himself that could and could not be crossed. Inevitably, he would somehow cross one of those boundaries and commit a “crime” even though there were no other individuals to be victimized. This could happen in a number of ways: He may engage in self-inflicted harm. While there are no other individuals, he is still trying to hurt a person. This, the attempt to harm a person, is generally viewed by most societies as a crime.

He might try to commit suicide. Suicide, by definition, is a homicide, and is recognized as a crime against society and against a person in many societies. Even if he is alone on the island, Durkheim (1895) would contend that the act of trying to kill oneself is more than a crime against religion or against self, but also against the society in which one lives.

He could commit any number of potential crimes against his newfound society that involved self harm, or other unspecified crimes which he might define for himself as a part of his society. But why would he do these things and what does it tell us about human behaviors and what does it tell us about the identification and prevention of victimless crimes?

It tells us something about human behaviors in that it would re-affirm two criminological theories and their suppositions about human nature. First it would lend credence to Beccaria (1764) and subsequent researchers’ claims that humans are born hedonistic. If we are all born with an inherent desire to maximize our pleasure and minimize our pain, why wouldn’t Crusoe do whatever he wanted on the island, even if it was against his own rules? He has nothing in place that will deter him from committing a “crime” against the island society he has built around himself; he is certain that there will be no punishment, therefore everything is fair game.

That would not, however, explain any sort of physical self harm or attempted harm. However, Reiss (1951) may have a solution to that and to the other deviant behaviors he might engage in against the island society. According to Reiss there are three components of control and deviance occurs when one of them breaks down: a lack of proper controls developed internally during childhood, a breakdown of these controls, or a lack or absence of societal rules and norms.

We have already established that Crusoe is in a state of normlessness, having been cast off from the society he has known and has been left to establish a new society based on his own beliefs. He faces a unique lack of societal rules and norms he was brought up with. He is no longer bonded to the rules of the society (Hirschi, 1969) that he has left behind and must forge ahead to create a new law unto himself.

It is therefore possible that he will use a set of neutralization techniques, as described by Sykes and Matza (1957) to free himself from the constraints of his old society. These could be in a number of techniques: Denial of responsibility, denial of injury, and misrepresentation of the consequences.
In the case of Crusoe, he could easily claim denial of responsibility. He has been forced into a situation far beyond his ability to control. Therefore, he must do what he has to in order to survive.

He could also easily claim denial of injury. He isn’t hurting anyone (except possibly himself), because there is no one on the island to hurt. If he sets a boundary for himself of how many coconuts a day he can eat, and decides to eat two more, he is being a deviant, possibly criminal. Durkheim would say that he’s engaging in the acts that form a natural boundary of society (1895) and Sykes and Matza (1957) would say that he’s justifying the violation of his self-imposed rules by denying an injury occurred.
Finally, there is the misrepresentation of the consequences. This is when the offender focuses mainly on the rewards and minimizes the consequences of his actions.

In the case of Crusoe, the coconut example may be a good one. If he has set a rule for himself on the number he can eat per day and violates that, he may deny the injury; but if he does this consistently, he may risk his food supply. He has then become too focused on the reward, and is ignoring the possibility that his crime against his own rules may result in dire consequences for the amount of food stores which he has access to on the island.

What Crusoe’s actions or lack of actions on the island show us about victimless crimes, is that it is easy for us to surrender to our more base instincts when we do not have a naturally defined boundary. It is easy to cross that boundary and enter into a new line of thinking if we are not contracted and bound properly to society.

It is easy for an individual, as Crusoe has demonstrated in these scenarios, to “cross the line” and go into deviant and sometimes criminal behavior without the proper deterrent or bond to society. In a modern setting, some “victimless” crimes might be prevented from making certain that people are bonded to society during their youthful years (Reiss, 1951, Hirschi, 1969) through programs that encourage community activities or positively reinforce the norms and mores of society. Others may argue that “victimless” crimes might be deterred by the use of more police officers, thus increasing the certainty of punishment (Durlauf & Nagin, 2011).

As a society, we like to think that only outsiders and loners are responsible for crime, but this is not the case. It is difficult to identify crime, especially victimless crimes, because it can at times be a private manner. Not everyone is going to be on the street corner smoking a joint. While Crusoe’s experience on the island can teach us much about how to recognize problems associated with victimless crimes and preventing them, it does not help us to identify how it will occur. Humans are social creatures, and one man alone on an island does not interact enough for us to gather sufficient data on something that could be entirely different in a social context where the rule of society and the interaction with others applies.

[MissyP89]

I prayed lots. Now pass.

[BG45]

Thank you. I appreciate that so very very very much. And I intend to. Hoping God's like "Here, worst subject is first. Go get it, it's downhill from there and you have 4 days until the next two start."

[BG45]

I've done more than i thought practice question wise. It's only been since Thursday since I sent my last batch to my reviewers. I just e-mailed 5 new practice answers.

[BG45]

[center][b](Again, apologies, PM hates my formatting from Word.)[/b][/center]
[center][b]January 2011[/b][/center]
[center][b]Research Methods Qualifying Exam[/b][/center]
[center][b]Morning Questions[/b][/center]

[b]Instructions:[/b] The morning session is comprised of two (2) questions. You have three (3) hours to answer the morning questions. Answer the questions asked. Read the entire question before you begin to write your answer. Each question is divided into several parts. Answer each part of the question and label your answers (1a, 1b, etc.).

[b]Question 1: General Research Methods (33% of grade)[/b]

Sex offender registration and notification laws have been enacted under the guise of providing protection to US citizens. It has been argued that if community members are notified of the sex offenders in their area, they will take measures to protect themselves and others from sexual victimization. You are now being asked to design a [b][i]quantitative survey[/i][/b] of the extent of the use of notification (i.e. Megan’s law websites) or measures a community has taken to protect themselves. The study may be either descriptive or explanatory in intent.

Cover the following areas in your answer:
[list=1]
[*]State a hypothesis about the extent of use of notification websites or measures of protection that can be tested by use of quantitative survey.
[/list][list=1]
[*]Conceptualization and Measurement:
[/list][list=1]
[*]Provide BOTH a conceptual or theoretical definition and an operational definition of either extent of use of notification websites or measures taken for protection.
[/list][list=1]
[*]Develop two (2) survey items, including response categories, to measure either extent of use of notification websites or measures taken for protection.
[/list][list=1]
[*]Explain how-in the context of this study-you plan to enhance reliability and validity of your measure.
[/list][list=1]
[*]Sampling: Your goal is to generalize your findings. Which sample design will you use? Justify your choice. Explain how subjects will be selected (that is, demonstrate that you understand how the sampling technique will work).
[/list][list=1]
[*]What limitations do you see in your design?
[/list]


[b]Question #2: Qualitative Methods (33% of grade)[/b]

After conducting the quantitative study on the public use and change of behavior based on sex offender registration and notification, you decide to delve more deeply into how the community views the registry, notification laws, and sex offenders. Assume that you have access to individuals who completed the quantitative survey and have agreed to participate in a more in depth qualitative study.
In your answer, be sure to address the following:
[list=1]
[*]State your research question(s)
[/list][list=1]
[*]Explain which qualitative method(s) you would use to answer your research question(s), and explain why it is superior to other qualitative approaches. [i]Be certain to provide enough detail to demonstrate that you understand the method you propose and[/i], [b]demonstrate[/b] that you know the central differences between the qualitative and quantitative paradigms.
[/list][list=1]
[*]Describe your sampling strategy – be specific
[/list][list=1]
[*]Discuss any human subject protection issues raised by your design, and how you would address these.
[/list][list=1]
[*]Briefly describe how you will analyze the data.
[/list]



[b]Answer:[/b]
[list=1]
[*][b]Hypothesis:[/b] The measures a community takes to protect itself from victimization are unrelated to victimization.
[*][b]Conceptualization and Measurement:[/b]
[b]2a.[/b] Conceptual definition for community protection measures:
Community protection measures consist of any efforts or measures that a community takes as a group to prevent any physical or emotional harm to its members.

Operational definition for community protection measures: Community protection measures can include any of the following: Neighborhood watch programs, neighborhood meetings, increased child supervision (on and off playgrounds) in the community. These measures may be both reactive or proactive.

[b]2b. Survey items:[/b]

[i]Directions: For the following questions, please circle the answer that best describes your response.[/i]
[/list][list=1]
[*]Does your community engage in Neighborhood Watch style programs? (This would be a program where community residents are involved in watching for suspicious people who do not normally belong in your area.)

Yes No Uncertain
[*]In the last four months, approximately how many times have you attended meetings with the other members of your community in regards to safety precautions?

0
1
[list=1]
2
[/list]
[/list]
3
4
[indent=1] 5 or more[/indent]

[b]2c. Enhancing the reliability and validity[/b]

Reliability and validity are very important to any study. Therefore it is important to lay the foundations as to how to properly measure them. The first type of reliability that will be utilized for this study will be that of test-retest reliability. A group will be given the survey instrument once and then again in four weeks period, as is recommended for the optimal time between the administrations of these instruments (Carmines & Zeller 1979; Maxfield & Babbie, 2008). This will allow the researcher to determine if what is being measured is being done in a reliable manner, meaning there is not major deviation from the previous results. If it is not, then there is test-retest reliability.

The most commonly used example for this is the bathroom scale. If you get on twice in a row, your weight should be the same. In this case, if someone takes the survey and would respond by saying they’ve attended one community meeting about safety in the last four months, it would be conceivable that this result would not jump to 5 or more in the space of only four weeks.

Also because of the nature of using scale items, a Cronbach’s Alpha will be used to make certain that the items “hold together” by correlating at a level of .70 or above (Nunally, 1979). This should aid in the reliability, by showing that these all tap into the same underlying construct.

Validity will be addressed in three ways. These will be face, content, and construct validity. For face validity, the researcher must determine that the survey instrument measures what it intends to on its face. The researcher looks through the questions and they do appear to measure the precautions taken by a community and the level of participation. Furthermore, it will be sent to experts in the field for review, such as Levenson, Tewksbury, and Sample; if they agree that it appears to measure the precautions, then the researcher feels confident that it is.

For content validity, the researcher addresses if the full content of the domain is being looked at. That is, the survey instrument will tap into the minor forms of protection such as watching one’s child or neighbor’s child while playing in the yards and street to rather major forms of community protection measures, such as the neighborhood watch programs.

Construct validity relates to the theoretically predicted relationship. In this case, the hypothesis is that the community protection measures and precautions will not be related to the community level of victimization. This will be assessed by comparing the observed rates of victimization via official data and the reported protection measures. The construct will hold strong, as previous research has shown that the rate of victimization is almost never tied to the actions of communities to protect themselves following a notification of a sex offender living amongst them (Bandy, 2011; Levenson, 2010; Sample, 2010; Zgoba, 2010).

Construct validity will be further addressed by running a factor analysis on the survey instrument. When the factor analysis is complete, it will show one underlying construct with an eigenvalue greater than one; this will show that the survey items are indeed tapping into the construct the researcher has designated as community protection measures.
[list=1]
[*][b]Sampling and Generalization[/b]
[/list]

The researcher has decided for this study to use the population of a nearby metropolitan area for the population (and sample). The Pittsburgh metropolitan area provides a diverse population in the areas of ethnicity, socio-economic status, and living arrangements (urban, suburban, and rural).

For this study the researcher has chosen to use a stratified random sampling method for this population. The researcher will stratify each of the cities in this metropolitan area. Then he will obtain a list of the addresses within each of these cities from the sister agency of the United States Postal Service which maintains addresses (sans names) for sale for research and marketing purposes (Dillman, 2009). Each of the addresses in the city will be placed in an order with no particular significance with a number assigned to each; this ensures that there is random selection based on the number, not a certain neighborhood within a city.

The researcher will then select one-third of the addresses in each city for inclusion in the survey. He will mail pre-notice letters asking for their participation and explaining what to expect from the next mailing, of the survey instrument and informed consent form. One week later, the researcher will send out the survey instruments via the mail with the instrument that asks questions about the community protection measures given.

This will ensure that there is an equal representation of every city within the metropolitan area and provide for an acceptable sample size.[list=1]
[*][b]Limitations[/b]
[/list]

The researcher acknowledges that there are certain limitations to this study. The first is that this is a mail related survey, and that means that there is a chance of a low response rate. The researcher has addressed this by sending out surveys to one-third of the potential population that is eligible, thus with even a 15-30% response rate (Dillman, 2009) there should be ample responses to reach proper statistical power to run the tests.

The researcher acknowledges that these results may be skewed due to only certain socio-economic classes of people or races being more likely to respond. In order to combat this potential threat, the researcher is including a self addressed and stamped envelope to return the completed survey. According to Dillman (2009), people are most likely to return a mail survey when they have to do the least amount of work to return it; this would be the least amount of work because it only requires being sealed and dropped in a post office mail slot or a mailbox on a street within their city.

There is also the limitation of monetary cost. Surveys in the mail are an expensive proposition, as the pre-notification letter, the printing and mailing of the surveys (with paid postage to mail back), a survey follow up for non-responders, and a thank you for those who have participated quickly adds up. This is especially true of the size of the sample at hand.

The final limitation of this survey is that it may yield results contrary to the desire of the expressed majority opinion. Community protection measures and SORN policies are often a feel good measure backed by no empirical evidence (Burchfield, 2011). Even if the expected hypothesis is proven correct and published in the field, it will never see the light of day in the general public due to going against the prevailing opinions and the media generated myths of sex offenders (Harris et al., 2010).



[b]Question Two: [/b][list=1]
[*][b]Research Questions:[/b]

A. What community protection measures does your community use?
[/list][list=1]
[*]What sort of threats or issues do you think your community requires protection from?
[/list][list=1]
[*]What are your thoughts and feelings on sex offender registries?
[/list]

D.Are sex offenders a problem or threat to your neighborhood?[list=1]
[*][b]Methods Used (and difference from Quantitative)[/b]
[/list]
There will be one primary method used for this study. The method that will be used is semi structured in depth interviews (Maxwell, 2005).

The semi-structured interview method will allow the researcher to ask questions and then follow up questions as well. This will also allow the researcher to change the structure and flow of the interview to suit the person being interviewed, rather than being locked into a rigid question and answer format that doesn’t allow for deviations. This flexibility also leads to greater respect and bond between researcher and subject.

These interviews will be recorded on the researcher’s digital recorder (with permission) in order to preserve the conversation for transcription later. In addition to recording, the researcher will be taking notes during the course of the interview; this is for two reasons: First, it enables the researcher to keep track of his thoughts when looking back at the interview later and second, because it acts as a safeguard should the recording be flawed or the sound muffled in some way. In an ideal world, the researcher will not be faced with these difficulties, but Murphy’s Law applies to research as much as any other field, and it is always best to be prepared for unexpected difficulties such as a battery in a recording device dying, or the sound not being captured adequately.

The researcher will not use focus groups ( a group interviewing strategy) for this research, because the subject matter does not lend to their nature. The nature of a focus group with its group interactions makes it difficult to maintain confidentiality as more people have to hold their silence about what occurs. Also, people who tend to talk more or have more vociferous opinions are most likely to dominate the group. This also doesn’t help for people to be asked follow up questions about a subject.

The nature of quantitative research is different from that of qualitative. In the quantitative paradigm, it is more about the numbers and straightforward questions about how many times someone has done something, did you do X, and items that can easily be put into a form where a statistical analysis may be run. From the qualitative perspective, it is more about asking the who, what, where, and why; qualitative research allows for an expansion with follow up questions and rich description of what has gone before.

The first question/preceding study was a quantitative study. It used a rigid survey instrument to ask questions that could be answered with numbers or a yes/no/unsure format. This study uses a qualitative format where the researcher is the instrument and the semi structured in depth interviews allow for the researcher to ask follow up questions and explore at a greater depth, the emotions and ideas behind the community protection and sex offender issues, rather than comparing official victimization rates to measures taken, as the quantitative sister study did.[list=1]
[*][b]Sampling strategy (be specific)[/b]
[/list]

The researcher will utilize a purposive random sampling strategy for this study (Creswell, 2007). Since this is a continuation of the previous study, it makes sense to use this. Rather than reinvent the wheel by drawing an entirely new sample, the researcher will use those surveys who responded that they do take precautions and community protection measures and place them in a list.

The researcher will then take this purposefully created list and contact one-third of the individuals on it by selecting every third address. The names will be unknown still, as the list of addresses related to the survey instruments does not have names, but only the street addresses. They will be contacted via mail and asked if they wish to participate in an interview regarding their feelings that they expressed on the instrument and given the researcher’s contact information again (though they may already have it from their copy of a debriefing statement sent with the informed consent from the first study).

It is hoped that this will result in a total of 25-50 residents who are willing to be interviewed, however, if saturation point is reached prior to this, the researcher will discontinue the interviews. If this amount is not reached and further interviewees are needed, the researcher will send out mailings to the next one-third of the list to be chosen by every third person who has not been previously selected.
The researcher hopes that this will provide for a sample which includes rural, urban, and suburban individuals of various races, ages, and socio-economic statuses.[list=1]
[*][b]Human Subject Protection Issues[/b]
[/list]

The researcher believes there will be three primarily issues that must be addressed: Anonymity, bias, and sensitive topics. The issue of anonymity, the interviewee’s name has not been given at any point prior to the researcher and interviewee meeting unless the interviewee has given it in contacting the researcher. In order to preserve this, the researcher will not ask the name of his subject, but merely introduce himself and ask if there is anything he or she wishes to be referred to as. If they give a real name, the researcher will smile and nod and pick another first name from the phone book or from a book of baby names. Care will be taken to describe the individuals with qualities instead of physical features to add a further layer of protection for their identities in the write up of the study (Maxwell, 2008).

The second issue is that of subject comfort. This should not overly impact the human subjects involved in this study, but the researcher will make certain to take efforts to minimize discomfort caused to them. The interviews will take place in a safe environment and water will be provided in case the subject would need to sip something to feel comfortable or to maintain their voice. This safe location should have comfortable chairs or other furniture so that the subject is able to feel more relaxed.

The third issue is that of the sensitive topic being addressed. These individuals have responded that they do take measures for their safety and the safety of their communities, and one of the research questions specifically involves sex offenders. This means that the interviews must be handled with the utmost care. Questions should be worded with care and to avoid leading in any particular direction if possible.[list=1]
[*][b]Analysis of Data[/b]
[/list]

As previously mentioned, the researcher will be recording the interviews via a digital recorder and taking notes during the course of these interviews. The notes will allow for the researcher to know what he was thinking during the course of the interviews and will serve as a back up in the event that the recording device fails to properly relay information. Also, these notes allow for things that a basic digital recorder does not, such as body language that has been used during the course of the interview.

After the data has been obtained, the researcher will place this into a Qualitative Research program such as Qualrus or Nvivo to code for analysis. These programs help a researcher identify trends or patterns in data that has been collected, but of course, will not replace the existence of a thinking researcher.

[BG45]

Some feedback I got back on the January Sex Offender question:
1) You're sucking today Phil
2) You should just be doing a descriptive, it doesn't really fit to try to force it into a regression model like you are.
3) Under the hypothesis: Maybe the sex offender registration notification system has not made a difference in the community awareness of sex offenders residing in the area. We're not talking about if victimized, if so the DV should be assaults. But that isn't what's being asked. What's being asked is the "extent of use", such as how often do you use the registry, do you know it exists, how often do you check it. I'm setting this up to be a regression base when all I need is descriptive frequency data.
4) Conceptual: very vague.
5) Operational: More specifics needed, how do these relate to sex offenders and the sex offender registry.
6) Survey items: Have them leave a number for last four months instead of giving numbers, no need to force ordinal data when the question requires a ratio level response. Again, how does this issue relate to the sex offender issue when tey could attend multiple meetings for property crime issues, drug issues, etc. Get specific.
7) Could ask: How many times in the past three months have you reviewed the online sex offender registry for your immediate area (two blocks in all directions)
8) Could give an example about "how much do you agree with the following questions" and ask something like "Since the sex offender registry started in our area, the community has increased its direct supervision of children under 17 years of age" and place a line to see from Strongly Disagree to Strongly Agree.
9) Under enhancing reliability and validity: Talking to them is likely to increase the probabilty of some sort of action I'm my own historical threat, so test-restest reliability is not good for this. perhaps a split halves method would be better? if we were to do test-retset, I would need to know who completed each one with some sort of control number. This would possibly be best by putting in a 3.5 card to mail back with name and address. Also to enhance rel/val, you could do a pilot study and have expert reviews. In the event of test-retest, same address does not necessarily equal the same person taking it. As a segue into the Qualitative, you could say that if you would like to, put your name and number if you would like to discuss further in person.

Question Two:
1) Why only those who take precautions? This inherently biases the findings more than Qualitative normally is. It would be easier to take just a 9 block area and tease out why a difference of opinion might exist. Remember that everything is a tradeoff in validities, having this huge a sample is great for external validity, but it is awful for internal validity. It is probably best for me to take a small neighborhood as a sample, because you can always sample more neighborhoods. It will be easier to root out maturation, historical, attrition, etc. impacts on a small scale.

2) Why not interview registered sex offenders? They love to talk, because they're social pariahs and rarely get to. They'd also offer a unique perspective.

[brandelynmarie]

[b]Assumptions of Social Control Theory[/b] – [i]Both social and self control. Assume people are naturally inclined to deviate but are constricted due to self control or bonds to society. [/i]


I have this theory, but I call it Original Sin(inclination to deviation) , Interior Grace (Holy Spirit within self) & God's 10 Commandments/Church Teachings (external/society). But this is just my spin on it. You are in my prayers BG!