How to Conduct a Survey - Georgia Department of Community Affairs
How to Conduct a Survey December 5, 2017 Michael Casper, Compliance Manager How to Conduct a Survey 2 Purpose of a survey is to document that the proposed Community Development Block Grant (CDBG) target area meets HUDs Low andModerate Income (LMI) National CDBG Objective Requirement. Specifically, the proposed target area must have, at minimum, 70% LMI.
How to Conduct a Survey 3 For Detailed information regarding How to Conduct a Survey and Survey requirements, see the 2018 Applicants Manual Appendix C Changes to Appendix C 4 1. See the Special NOTE for Revitalization Area Strategy (RAS) surveys. ---- Brief Summary ---- Typically, the American Community Survey
(ACS) data has been used to determine the number of households in Poverty [census block group(s) of >20%]. However, the ACS may have a high error margin. An alternative is to survey the ENTIRE block group(s). Changes to Appendix C 5 2. Random Sampling When conducting less than a 100% target area survey (to determine LMI percentage), a random sampling technique must be used.
The purpose of the random sampling technique is to ensure everyone who is in the group (CDBG target area to be surveyed) has an equal chance of being included in the sample Changes to Appendix C 6 2. Random Sampling (continued) CDBG or RAS applicants, when conducting less than a 100% survey, MUST use a Random Number Generator in
conducting a random survey Changes to Appendix C 7 2. Random Sampling (continued) [NOTE: Random Number Generators are available free on-line] How to use a Random Number Generator a. Assign sequential numbers to your Universe (addresses in the proposed CDBG target area) (Also, see Defining Your Universe section in Appendix C) For Example: Assigned # a.
b. c. d. e. f. g. h. i. j. 1 2 3 4 5 6 7 8 9
10 123 125 127 129 131 133 135 137 139 141 Address___ Happy Lane Happy Lane Happy Lane Happy Lane
Happy Lane Smiley Place Smiley Place Smiley Place Smiley Place Smiley Place Changes to Appendix C 8 2. Random Sampling (continued) b. Enter Universe size Number Generator
c. Enter the desired number of random numbers to generate (as determined by the sample size and expected response rate calculations See Appendix C) d. The resulting listing is then assigned to the sequential numbering in (a), above. into the Random Changes to Appendix C
9 http://stattrek.com/statistics/random-numbergenerator.aspx Changes to Appendix C 10 2. Random Sampling (continued) Based on the Random Number Generator, and using the Universe example, we would select: Assigned # a. b. c. d. e.
f. g. h. i. j. 1 2 3 4 5 6 7 8 9 10 Address___
123 Happy Lane 125 Happy Lane 127 Happy Lane 129 Happy Lane 131 Happy Lane 133 Smiley Place 135 Smiley Place 137 Smiley Place 139 Smiley Place 141 Smiley Place How to Conduct a Survey 11 Important Considerations for Interviewers Interviewers should
attempt to contact respondents at a time when most likely to get high response rate Interviewer should ensure each respondent is asked the exact questions as other respondents Interviewers should ensure questions are not biased or loaded questions should not imply that lower reported incomes equals higher probability of area receiving CDBG funding How to Conduct a Survey
12 The following slides provide more survey methodology information and detail Low and Moderate Income (LMI) Benefit 13 I. Documenting LMI Benefit (i.e., meeting a National CDBG Objective requirement) for the CDBG project: A. Direct Count B.
Area Benefit 1. II. Area Benefit Surveys DCA Form 6 Instructions I. Documenting LMI Benefit 14 Two Methods A. Direct count based on client records Housing Job Creation such as Employment Incentive
Program (EIP), etc. Limited Clientele for Community Services Buildings B. Health Centers, Senior Centers, etc. Area Benefit Public Infrastructure and other Area Benefit projects I. Documenting LMI Benefit Contd 15 I. A. Direct Count - Limited Clientele Benefit (LMC) Some clientele can be assumed to be LMI
Only need a count of the # of people Those Assumed LMI (Not for Area benefit): Elderly Severely Disabled Homeless Battered or Abused Men, Women or Children Migrant Workers Persons living with AIDS Illiterate I. Documenting LMI Benefit Continued 16 I. A. Direct Count - Limited Clientele Benefit (LMC) Contd
If clients are not assumed to be LMI (e.g., Health facility clientele, etc.), the documentation of the number of persons to benefit must include family size and income data 17 I. Documenting LMI Benefit Continued 1. B. Area Benefit Surveys Area Benefit projects include: Water and sewer projects Streets and drainage projects, etc.
Count everyone in area to benefit All residents on a street Separate target areasNOTE: One very low income area can not qualify a non-low and moderate income area - even if overall benefit exceeds 70% minimum 18 I. Documenting LMI Benefit Continued 1. B. Area Benefit Surveys continued When properly conducted, Area Benefit Surveys Can be Accurate
Refer to the Applicants Manual Appendix C, to determine when a 100% survey is required versus a Sample Survey Prefer 100% survey for a small area Large area may require a sample survey, such as a: Water Storage Facility Treatment Facility 19 I. Documenting LMI Benefit Continued
1. B. Area Benefit Surveys Continued The goal of a sample survey is to make an accurate inference about a population based on a survey of a smaller or sample group Sample surveys can be, generally, accurate if properly conducted 20 I. Documenting LMI Benefit Continued 1. B. Area Benefit Surveys - Continued
Sample Standards Must be large enough - determination of sample size is based on population Table 2 (located in the Applicants Manual Appendix C) gives required minimum sample size based on population or universe Example: 100 to 115 households requires a sample of at least 90 households 651 to 1,200 households requires a minimum sample of 300 households 21 I. Documenting LMI Benefit Continued 1. B. Area Benefit Surveys - Continued
Must be a systematic and random selection (using a random number generator) of families to be surveyed so that each family has an equal chance of being selected Going door-to-door (e.g., every other address) until you accumulate enough surveys to meet the minimum sample size is not random. Guide provides more information regarding selection methods (See Applicants Manual, Appendix C) II. DCA Form 6 Instructions 22
See Applicants Manual, page 65 Must describe detail on how the information was determined (survey methodology). If a survey (100% or random) is used, DCA-6 must include:
Description of how the # of households was determined Description of how sample was selected Describe a random sample method (if applicable) See Appendix C for guidance Description of how the survey was conducted (Who and When) Provide a copy of the actual survey form A table of the results is helpful Prior to funding decisions we may ask for copies of all surveys completed Must keep all information for DCA review II. DCA Form 6 Instructions - Continued 23 Basic data that must be included on DCA-6: # of households in area
# of households interviewed (Note: a vacant house is not a household) # of LMI households interviewed # of people living in LMI households # of above LMI households interviewed # of people living in above LMI households Remember to use the LMI worksheet (Appendix C page 10) for sample sizes of less than 100 percent. II. DCA Form 6 Instructions - Continued 24 Example Survey DCA-6
II. DCA Form 6 Instructions - Continued 25 Example Survey Methodology Description II. DCA Form 6 Instructions - Continued 26 Example - Survey Summary 27 For community-wide projects:
HUD Website for LMI estimates for all cities and counties. https://www.hudexchange.info/programs/acs-low-mo d-summary-data/acs-low-mod-summary-data-blockgroups-places/ Questions? 28 Michael Casper, CDBG Compliance Manager (404)679-0594 or [email protected]
Decision-Matrix (Pugh's method) A weighted decision matrix Design Selection -Pugh's Method The method is an iterative evaluation that quickly identifies the strongest design solution. Step 1 -Select the Criteria for Comparison The list of criteria must be developed from the...
Interact (Battle Ground, Union) High school key clubs (Union, Mountain View) Green Teams (Cascade Middle School, Educational Opportunities for Children and Families) King Middle School Drumline. And we partner with the Northwest Wild Fish Rescue. We provide them with volunteers...
Converting SI units is just as easy as converting pennies to dollars. KHDMDCM Killing Horses Dead Makes Davey Crockett Mad Solving Problems Solving Problems Convert 142 km to m Looking for: …the distance in meters Given: …distance = 142 kilometers...
Who lyved euere in swich delit o day That hym ne moeved outher conscience, Or ire, or talent, or som kynnes affray, Envye, or pride, or passion, or offence? I ne seye but for this ende this sentence, That litel...
Your Name The Company you represent Where you operate Where We Are Now About the UKLCN Background - born in 2003 Dormant - but the site chugs away Revival process - establish priorities Establish existing Membership Target new Members Meet...