Presentation is loading. Please wait.

Presentation is loading. Please wait.

Presentation of Results Mr. Joel LaLone Joel LaLone Consulting Watertown, New York April 11, 2014.

Similar presentations


Presentation on theme: "Presentation of Results Mr. Joel LaLone Joel LaLone Consulting Watertown, New York April 11, 2014."— Presentation transcript:

1 Presentation of Results Mr. Joel LaLone Joel LaLone Consulting Watertown, New York April 11, 2014

2 Overview/Goals of the Study Collect tobacco-related information (attitudes, opinions, and behaviors) from current adult residents of Oswego County to: 1.Develop a greater understanding of current attitudes, opinions, and behaviors regarding tobacco in Oswego County. 2.Use this data to make data-driven decisions in the development and planning of future TFNOC programming and activities. 3.Use current data and past data to identify trends and evaluate impact.

3 Methodology →Telephone survey, instrument developed by NYSDOH, TFNOC, and JLC. → ≈30 survey questions ( ≈25 tobacco-related, 5 demographics) →Calls made between 3:30-9:00 p.m., M-F, January 2014 →Random residential landline and cellular numbers (≈24%, 20% c-o) →All calls made by JLC employees trained in Human Subject Research laws and effective interviewing techniques, from a call center in Watertown, NY →400 adult participants (18+ years of age) →≈40% response rate → At least 4 call backs to each no answer/busy → Margin of Error: →approximately ±3.9% →Margin of Error is larger than ±3.9% when investigating smaller subgroups (n<400) →Telephone interviewing historically selected as the sampling protocol since it is far less susceptible to bias, and often-times less expensive, than: Mail surveys; Intercept (face-to-face) surveys; Focus groups; or Online surveying Sample Size (n=…) Approximate Margin of Error 30 ±14.3% 50 ±11.1% 75 ±9.0% 100 ±7.8% 125 ±7.0% 150 ±6.4% 175 ±5.9% 200 ±5.5% 225 ±5.2% 250 ±5.0% 275 ±4.7% 300 ±4.5% 325 ±4.3% 350 ±4.2% 375 ±4.0% 400 ±3.9% Margins of Error for Varying Sample Sizes

4 More Methodology – Error Before we begin the roll-out of highlighted results in 2014, it is predicted that each of you … at various times … for varying survey results ….. during this presentation … will experience …  Surprise, and  Confirmation, even seemingly obvious confirmation, and  Disagreement, and even possibly  Disbelief (and, for three of these four reactions, the natural response is for one to question the methodology in the study!)

5 More Methodology – Error There are three, and only three, sources of error when completing research based upon sampling (not exclusive to survey research) RANDOM ERROR chance of sampling (how can n=400, out of N=90,000 in Oswego County, be believed? Sampling -- Coins? Aspirin? Nate Silver …. 1 out of 2,252,000,000,000,000) SAMPLING ERROR the way the sample was selected (under-represented groups …. Quotas vs. Weighting … school employees … cell-only persons … JLC weights by SEX, AGE, ED, HH TYPE, and PHONE OWNERSHIP) NON-SAMPLING ERROR what was actually measured (broken thermometer? … definitions, phrasing, non-leading, consistency, and constant supervision)

6 Final Methodology – Controlling Error Confidence Level and Margin of Error As a result of using n=400, and weighting the data to adjust for sampling error, and using IRB approved valid survey questions with well-trained interviewers -- We are 95% confident that a sampled % that is reported in this study for Oswego County is within approximately ±4% of what would be found if, in fact, all adults in the county were surveyed.

7 Presentation of Results The interviews included eight sections of tobacco-related survey question, all of course, tied to the TFNOC workplan. Survey Sections: 1.Spreading the Message about the Dangers of Tobacco 2.Tobacco Marketing – Displays in Stores 3.Tobacco Marketing – Advertisements 4.Tobacco Point of Sale 5.Outdoor Tobacco Policies 6.Smoke Free Workplace 7.Smoke Free Housing 8.Tobacco Use

8 Presentation of Results – Highlight #1 (Table 22, Page 47) What more may we learn from the “not allow at all” statistic (36%)? – “framing” that statistic…

9 Presentation of Results – “Framing a Statistic” With survey-generated data, often times the following five methods are used to better understand the presented statistic – “framing” that statistic:  Within Response Distribution (Majority? 4:1?)  Trend Across Time (Increased? Decreased?)  Compare to Target/Benchmark (compare to current 21-county regional average? Workplan goal?)  Ranking among similar variables (First? Last?)  Cross-tabulations by Potential Explanatory Variables (Smokers and Non-smokers differ? Age-dependent? Gender- dependent?)

10 Presentation of Results – A complete analysis of opinions about not allowing smoking in residential units in MUDs in Oswego County. (page 47 in final report)

11 Presentation of Results – A complete analysis of opinions about not allowing smoking in residential units in MUDs in Oswego County (graphical)

12 Presentation of Results – Incredibly large number of options … so Abby had me focus on…

13 Presentation of Results – Highlight #2 (page 41 in final report)

14 Presentation of Results – Highlight #3 (page 42 in final report)

15 Presentation of Results – Highlight #4 (page 28 in final report)

16 Presentation of Results – Highlight #5 (page 32 in final report)

17 Presentation of Results – Highlight #6 (page 33 in final report)

18 Presentation of Results – Highlight #7 (page 34 in final report)

19 Presentation of Results – Highlight #8 (page 38 in final report)

20 Presentation of Results – Highlight #9 (page 55 in final report)

21 Presentation of Results – Highlight #10 (to make it an even “Top 10”) (page 30 in final report)

22 Any other survey results of interest? We can start digging deeper into the study results …

23 If you have further technical or statistical questions… Joel LaLone joel@joellaloneconsulting.com (315)-408-9214


Download ppt "Presentation of Results Mr. Joel LaLone Joel LaLone Consulting Watertown, New York April 11, 2014."

Similar presentations


Ads by Google