Attachment G - Assumptions for Calculation of Burden Estimates

Assumptions for Calculation of Burden Estimates in Table A.4

The following assumptions were used for the calculation of burden estimates in Table A.4.

1. DCI

• Telephone interview with grantee staff on program design (Instrument DCI-1) (burden for program staff). We will conduct one-time, one-hour interviews with one staff member from each of the 60 stage-two grantees. The annualized burden is thus 20 hours in each of three years (60 burden hours divided by 3 years).

• Telephone or in-person interviews with grantee staff on program implementation (Instrument DCI-2) (burden for program staff). We will conduct one-time, one-hour interviews, with up to 15 staff members from each of the 20 stage-four grantees (300 burden hours). The annualized burden is thus 100 hours in each of three years (300/3).

• Program participant focus group (Instrument DCI-3) (burden for program participants). We will conduct four focus groups at each of the 20 stage-four grantees, with up to 10 individuals per focus group (4*20*10=800/3 = 267 respondents each year). Focus groups will last an estimated 1½ hours each (800*1.5=1,200 total burden hours), so the annualized burden is 400 hours in each of three years.

2. DCS

• Applicant Characteristics (Instrument DCS-1) (burden for program applicants). Based on the average number of individuals enrolled each year by the 120 current HMRF grantees, we assume an enrollment of 16,500 fathers per year in RF programs (275 fathers per grantee across 60 RF grantees) and an enrollment of 84,000 individuals per year in HM programs (we have assumed that half of the programs will serve couples [300 couples or 600 individuals per grantee] and half will serve youth [2,200 youth per grantee] across 60 HM grantees). We have also assumed that 35 grantees will enroll a comparison group for the evaluation (11 grantees enrolling an additional 275 fathers each [3,025 individuals], 12 grantees enrolling an additional 300 couples each [7,200 individuals], and 12 grantees enrolling an additional 2,200 youth [26,400 individuals]). Assuming it will take 15 minutes (0.25 hours) for each individual applicant (father, individual, or member of a couple) to complete the applicant characteristics instrument, the total annual burden for applicants to enroll into HMRF programs is 34,281 hours.

• Applicant Characteristics (burden for program staff). We assume that 3 staff per grantee will conduct intake (360 staff total), and with a total of 137,125 intakes annually, each staff member will conduct an average of 381 intakes annually. Assuming it will take 6 minutes (0.1 hours) to enroll each individual applicant (father, individual, or member of a couple), such as screening for intimate partner violence and collecting contact information, the total annual burden for program staff to enroll individuals into HMRF programs is 13,716 hours.

• Program Operations (Instrument DCS-2) (burden for program staff). Grantees will be required to report performance in three operational areas: (1) marketing/outreach/recruitment; (2) quality assurance and monitoring (staff qualifications, staff training, staff supervision, and direct observations of workshop facilitation); and (3) implementation challenges. We assume that one data entry worker per grantee will enter these data, and that it will take 15 minutes for the first data entry and 10 minutes each quarter thereafter, for a total of 45 minutes (0.75 hours) per year per grantee. Thus, the total annual burden is estimated to be 90 hours.

• Service Delivery Data (Instrument DCS-3) (burden for program staff). Staff from all 120 grantees will need to document service delivery for performance reporting and for cross-site analyses. We assume that 60 percent of those in RF programs (16,500 fathers being served in the program multiplied by 0.60 = 9,900) and 85 percent of those in HM programs (84,000 individuals multiplied by 0.85 = 71,400) participate in some program service for a total of 81,300 individuals (estimates of participation informed by preliminary data from the Parents and Children Together (PACT) evaluation of HM and RF programs).

For each individual, we assume weekly service data entry for up to 12 weeks (likely the maximum length of a workshop), with quarterly service updates thereafter, for a total of 15 data entries per participant annually. Assuming each data entry will take 2 minutes (0.033 hours), it will take 0.50 hours (15 x 0.033 hours) for each of the 120 grantees to enter each participant’s service data annually. We assume that one data entry worker per grantee will enter these data, averaging 678 entries per worker. Thus, the total annual burden for entering individual-level service delivery data is 40,680 hours across the estimated 120 grantees.

• Pre-test/Post-test (Instrument DCS-4) (burden for program participants). Every participant will self-administer a pre-test and a post-test during the first and last workshop session (or one month after the pretest for programs shorter than one month), respectively. We have assumed that both those who receive services and those in the comparison group complete the pre-test and post-test. Based on participation rates of program group fathers and couples enrolled in the PACT evaluation, we assume 60 percent of RF program applicants (includes 16,500 fathers receiving services and 3,025 in the comparison group for an initial total of 19,525*.60 = 11,715) and 85 percent of HM program applicants (on average; includes 84,000 individuals receiving services and 33,600 in the comparison group for an initial total of 117,600*.85=99.960)—or a total of 111,675 individuals—will complete the pretest. We also assume that 40 percent of RF program applicants and comparison group members (19,525*.40=7,810), and 70 percent of HM program applicants and comparison group members (117,600*.70=82,320)—or a total of 90,130 individuals—will complete the post-test. This results in a total of 201,805 administrations. We assume each pre- and post-test will take 25 minutes (0.42 hours), for a total estimated burden of 84,758 hours (46,904 hours for pre-tests, and 37,855 hours for post-tests).

• Pre-test/Post-test (Instruments DCS-4HM and DCS-4RF) (burden for program staff). We have assumed that a small number of grantees (6) will not be able to administer the applicant characteristics instrument, pretest, or post-test using nFORM. Using the same rates of completion as described above, this will result in data entry for 1,953 RF respondents ((3 grantees/60 grantees)*(19,525 completing applicant characteristics + 19,525 * 0.60 completing pre-test + 19,525 * 0.40 completing post-test)=1,953) and 14,994 HM respondents ((3 grantees/60 grantees)*(117,600 applicants + 117,600 * 0.85 completing pre-test + 117,600 * 0.70 completing post-test) or 16,947 total. Those grantees will need to collect the data on paper and then a staff person will enter it manually into nFORM. We have assumed that this process would take 6 minutes per instrument (applicant characteristics, pre-test, and post-test) and that each grantee would have two data entry staff (each staff entering data on 1,412 participants or 16,947/12). The estimated total annual burden is 5,083 hours.

• Semi-annual Program Performance Report (PPR) (Instruments DCS-5HM and DCS-5RF) (burden for program staff). Grantees will also be required to report on a subset of the DCS performance measures twice per year for the semi-annual submission of their PPRs, as well as provide narrative updates on activities, accomplishments, and challenges. The quantitative summary of performance measures will be generated by nFORM. We assume that one data entry worker per grantee will require an average of 3 hours twice a year to prepare the semi-annual report. The total estimated burden is thus 720 hours.

• Quarterly Performance Report (Instruments DCS-6HM and DCS-6RF) (burden for program staff). Grantees will be required to report on a small subset of the DCS performance measures twice per year. The quantitative summary of performance measures will be generated by nFORM. Grantees will also need to provide a narrative on activities and context for the performance measures. We assume that one data entry worker per grantee will require an average of one hour twice a year to prepare the semi-annual report. The total estimated burden is thus 240 hours.

1