Firearm Storage Safety: An Analysis of Home Owners' Practices, Exams of Survey Sampling Techniques

Download Firearm Storage Safety: An Analysis of Home Owners' Practices and more Exams Survey Sampling Techniques in PDF only on Docsity! The County Board and Health and Human Services Office (HHSO) commissioned a study regarding the storage of firearms in the home. The County Sheriff’s office reported that there were 55,636 households in the county with at least one legally registered firearm. A list of un-attributed addresses for gun owners was obtained and organized by zip code. The addresses were then regrouped into three strata: City, Suburb, and Rural. A simple random sample was then taken from each of the three strata. The target population for this study is legally registered firearm owners that store their firearms in their home. A probability of selection of 1 in 105 was chosen to yield a total sample size of 516 households. The breakdown by strata is as follows: City (439), Suburb (51), Rural (26). The last group does not have sufficient sample size so either it had to be increased or you would obtain an SRS from 3 strata combined. Moreover, in your analysis you do not account for different strata so for consistency you could use location as a possible predictor if you think that it may be a significant predictor. As you state it initially you don’t have a nice representation of all three strata in your sample, when you adjust for non-response then you end up with even worse sizes. Each of the selected households was mailed a survey with a self addressed, postage paid return envelope. A total of 410 surveys were returned, resulting in a 79% overall response rate. Twenty-nine (7%) of the returned surveys indicated that their firearms were stored somewhere outside the home. So these surveys were not included in the data analysis. The County Board and HHSO are interested in determining if in general, firearms stored in the home are stored safely. The sponsors of this study also requested information regarding whether the presence of children and / or marital status affects the way in which registered gun owners store their firearms in the home. In addition to the specific criteria laid out by the sponsor, this study will also address the average number of firearms kept by gun owners, the purpose for owning the firearms, and the gender of the owners. The respondents consisted of 49 female primary gun owners and 361 male primary gun owners representing 12 % +/- 1.6% and 88% +/- 1.6% of the respondents, respectively. The majority of respondents, 63%, cited hunting as the reason they owned firearms; the next most cited reason, 34%, was for protection; followed by hobby, 1.7%, and professional requirement, 1.7%. The average number of guns per household was 2.72 +/- 0.218 with an average safety score of 5.07 +/- 0.143. Need to explain how this safety score is obtained, what is the possible range for it, otherwise a mean number of 5.07 does not say anything. Additional class proportions and their associated average safety scores are provided in Table 1. You stated that your target population consists of people who store guns inside the home but your output below provides statistics for “Gun stored outside the home”, which contradicts to you claim in the 1st paragraph. STAT 804 Final Report 11 DEC 2007 Formatted: Superscript Overall Proportion in Sample Std Error Avg Safety Score Std Error Female Gun Owners 0.12 0.017 5.55 0.437 Male Gun Owners 0.88 0.017 5.20 0.148 Marital Status of Households with Guns Married Households 0.46 0.026 5.02 0.218 Unmarried Households 0.54 0.026 5.20 0.182 Kids vs No Kids No Kids 0.50 0.026 5.25 0.198 Kids 0.50 0.026 5.23 0.200 Storage Location of Firearms Gun stored outside Hm 0.07 0.013 -- -- In Home 0.93 0.013 5.07 0.143 TABLE 1: Summary of selected proportions and average safety scores. Statistical analysis software, SAS 9.1, was used to test whether or not the level of safety used in storing a firearm is linearly dependent upon the marital status and number of children or whether there are any versus none? children in the home, children in the home only, and marital status only. Marital status was treated as a continuous variable, can not do that! having four possible outcomes: married, divorced, living with someone, and single. The effect Children have on safe storage of a firearm was tested by treating children in the home as a continuous variable, with four possible outcomes: none, one, two, three or more children in the home. If you want to treat it as a continuous variable then the actual number of children (not more than 3) should be used. The safety score was also a continuous variable made up by the product of the responses to two discrete variable questions. The first question, “What condition is the firearm stored?” The two possible outcomes to this question are loaded (score = 2) or unloaded (score = 1). The second question regarding safety, “Is your firearm secured when being stored?” The two possible outcomes to this question were, the firearm is locked in a gun safe (or disabled with a trigger lock)(score = 1), and the firearm left unlocked (or in a closet)(score = 4). Therefore, the safety score becomes a continuous variable with four possible outcomes, two which are considered safe and two that are considered unsafe. The safest outcome has value 1 and corresponds to the firearm being stored unloaded and locked in a gun safe (or disabled). The second safe outcome has value 2 and corresponds to the firearm being stored loaded and locked in a gun safe (or disabled). As you define it you just utilize the second variable “locked”, if it is then it is a safe way then why do you need the 1st variable if it does not affect your decision regarding safety. I guess I liked much more a coded variable that you proposed at the beginning of the term: consider multiple categories and assign values to all of them and then add to obtain a composite score. Then such a variable would be a much better continuous variable. The score you are currently considering is a very “artificial” variable The most dangerous method of storing a firearm is loaded and not secured; this outcome has value 8 in the model. Finally, an unloaded firearm that is left unsecured is also considered to be unsafe and has value 4. The generic form of the models tested is presented below: Z = a1X + a2Y + a0 Where X = Marital Status; Y = Number of Children; Z = Safety Score Formatted: Underline Formatted: Superscript