1999 National Household Survey on Drug Abuse Data Collection Report

3. DATA COLLECTION STAFF RECRUITMENT

The magnitude of the 1999 NHSDA required a field data collection management structure robust enough to support the interviewing staff and flexible enough to manage an ever-changing variety of issues. This chapter discusses the process of recruiting the staff needed to conduct the 1999 NHSDA data collection effort.

3.1 Regional Directors

The basic management structure used in previous NHSDA efforts included Field Supervisors managing states and substate regions and reporting to Regional Supervisors. Due to the size of the supervisory structure for 1999, it was impossible for the National Field Director to directly supervise the Regional Supervisors. A new level of management, the Regional Director, was created to serve in the position between the Field Director and the Regional Supervisors. Regional Directors (RDs) managed data collection within defined territories of the nation. Reporting directly to the National Field Director, the RDs, working with the Project Director and the National Field Director, served as the management team for all data collection operations.

The RDs were survey managers with many years of experience at RTI. Two of the seven RDs who worked on the 1999 NHSDA were Regional Supervisors on previous NHDSAs who were promoted to the RD position. The other RDs were staff from the Survey Research Division who had experience managing other large and complex projects.

Initially, plans called for four RDs—two starting in April 1998 to plan and manage the Counting and Listing operation, with two additional RDs starting later in the spring of 1998. These four RDs served during the development of field procedures and the recruiting and training of the initial Regional Supervisors and Field Supervisors.

After data collection began in January, 1999, managing the RD territories and providing other project management assistance was continually overloading the RD team. At the beginning of Quarter 2, two additional RDs were identified and the four RD territories were reconfigured into six to allow more focused attention on areas where production and staffing issues were still creating problems.

Each of the RDs managed a staff of Regional Supervisors (RSs), who in turn managed a staff of three to six Field Supervisors (FSs) who managed interviewers in their individual states or assigned areas. Each RD also managed a small staff of survey specialists at RTI who assisted the RD in a variety of functions, including monitoring various reports and measures of production and quality, and maintaining spreadsheets to monitor costs.

RDs also had project-wide ancillary functions not specific to their region. These included coordinating FS and FI recruiting; Counting and Listing activities; training activities; interactions with the Headway staff; and the supplying of material, equipment, and training packages to the field staff. The survey specialists assigned to the RDs assisted in these functional areas as well.

3.2 Regional Supervisors

Regional Supervisors (RSs) were the direct managers of three to six Field Supervisors. Reporting to an RD, RSs were responsible for all data collection activities in the state or states in their region. Initially, there were 18 RS positions on the supervisory team. Each of the eight large states was supervised by a single RS. The 43 smaller states, including the District of Columbia, each managed by one FS, were clustered geographically to be managed by the remaining 10 RSs. During the course of the year assignments were adjusted as RSs left the project team and new RSs joined; other changes allowed stronger, more experienced staff to manage troubled areas.

The Regional Supervisors were recruited from staff in the Survey Research Division at RTI, from staff who had previously worked on NHSDA as Regional Supervisors, and from a group of highly successful Field Supervisors. Initially, seven of the eighteen RSs were full-time RTI employees and the remaining 11 were previous field staff who became off-site, full-time RTI employees when they moved into the RS position. Working out of their home offices, these 11 RSs were provided with the same communications and computer capacities as the on-site RTI employees serving in the RS role.

3.3 Field Supervisors

Field Supervisors were the first-level supervisors of the interviewers conducting the data collection in each of the states. The FSs assigned work, monitored progress, resolved problems, and managed the day-to-day activities of the interviewers. Each FS reported directly to an RS.

Initial plans utilized 76 FSs—23 supervisors were on board by May, 1998, to supervise the listers completing C/L activities. An additional 53 FSs were added during the summer of 1998 to prepare for recruiting and hiring the interviewing staff.

Both groups of FSs were recruited using the same procedures. They were selected from a pool of over 200 candidates comprised of:

• FSs currently working on the 1998 NHSDA;



• FSs currently working on other RTI survey projects;



• FSs in the National Interviewer File (NIF) who had previously worked for RTI;



• Individuals referred by current NHSDA staff;



• Individuals identified by networking with other survey research organizations; and



• Individuals who responded to newspaper classified ads.

Each candidate was screened for interest and basic qualifications for the position. Candidates who successfully completed this initial screening were interviewed and evaluated by two or more of the RSs. A subset of the RSs conducted all of the interviews and reference checks using standardized materials.

After the interviews were completed, the interview summary and the reference checks for each of the candidates were compiled and distributed to the managers: RDs, the National Field Director, the Associate Project Director, and the Project Director. After reviewing the evaluations and references, the managers held an all-day meeting to discuss each of the candidates. Many of the candidates had worked on previous rounds of the NHSDA or other RTI projects and were well known to the reviewers. These qualified staff were placed in the pool of candidates to be offered positions. As other candidates were reviewed, they were placed in one of three groups: those to be offered positions, those needing additional review, or those who were not qualified. After further review, the members of the middle group were placed in one of the two other groups.

The initial group of 23 FSs were assigned territories generally based on geography since they managed C/L operations in states near their homes. In mid-summer, as additional FS staff were being hired for the screening and interviewing phase, all C/L FSs were contacted to determine their interest in remaining an FS (several listing FSs declined the opportunity to move to the next phase).

After identifying new candidates and continuing staff, the managers determined the placement of each potential FS by state. An effort was made to have the FSs manage states or territories close to where they lived. However, previous experience and matching skills of the FS with special needs in a state or area were the final determining factors in the placement of the FSs. The same team of managers then determined pay rates for the candidates based on experience and location of residence.

The RDs divided the pool of candidates and made job offers, discussing the details of the job offer and the pay rate being offered. The specific geographic assignments for each candidate were tentative until all offers were made and accepted. When several offers were declined, a few final shifts in territories were required.

Throughout the data collection efforts, the recruiting of candidates for FS positions continued. Each recommended candidate was screened for interest and basic qualifications. The successfully screened candidates were then interviewed by RSs and the summary of the interviews and reference checks evaluated by the RDs. Based on the evaluations, candidates were either placed in the "bullpen" or told that they would not be considered further. Candidates in the bullpen were used to fill openings that occurred when FSs left the project for personal or professional reasons or were terminated for performance issues.

During the year, there was a need to have a group of special FSs ("Super" FSs) available to cover FS territories when the primary FS was on vacation, to assist with special problem resolution, and to assist in the FI recruiting functions. With input from their RSs, each of the RDs reviewed the FSs in their regions to identify candidates for the Super FS (SFS) positions. In all of the regions except one, an FS was promoted to the SFS position and a replacement FS was hired from the bullpen. In one region, an SFS was recruited directly from the bullpen.

3.4 Field Interviewers and Traveling Field Interviewers

One of the primary FS functions was the continuous recruiting and hiring of the FI staff needed to complete the data collection work each quarter. The initial hiring target for the beginning of the first quarter was 1,200 interviewers. Approximately 200 of the veteran FIs working on the 1998 NHSDA were expected to continue, leaving a balance of 1,000 new interviewers to be hired for the 1999 NHSDA.

Interviewers who worked on the 1998 NHSDA with satisfactory performance ratings and who were located near sample locations were retained for the 1999 survey. Those interviewers provided the survey with an initial core of experienced staff.

Because of the greatly enlarged sample for the 1999 NHSDA, all FSs had to recruit new interviewers to staff their assigned territory. Some FSs only needed to supplement their staff because there were good performing 1998 NHSDA interviewers in their area. Other FSs whose territory did not include any 1998 NHSDA areas had to recruit their entire interviewing staff.

FSs used multiple recruiting approaches to identify candidates, including:

• identifying interviewers who worked on previous NHSDA surveys;

• reviewing the National Interviewer File that lists interviewers who have worked for RTI during the past 10 years;

• networking;

• placing newspaper advertisements and posting informational job flyers;

• contacting job service agencies; and

• using Internet job advertising and search services.

Networking involved any or all of the following contacts:

• other Field Supervisors;

• RTI staff working on other surveys with potential FIs available;

• other survey research organizations; and

• other Field Interviewers (current NHSDA FIs recommending successful candidates received a recruiting bonus).

A competitive hourly wage was offered to attract a large pool of candidates. Those with general interviewing experience, and especially those with experience working on government surveys, were given preference in hiring. However, candidates with transferable skills and experience—such as contact with the public, attention to detail, and organizational skills—were considered.

The work of an interviewer requires a wide range of skills and abilities. Some of the characteristics/qualities FSs tried to identify in potential hires included:

• intelligence;

• dependability;

• sensitivity/objectivity;

• voice quality;

• reading ability;

• listening skills;

• motivation;

• availability; and

• flexibility.

It was essential that staff hired to serve as interviewers understood and were committed to the standards of confidentiality and excellence required by the NHSDA. To help ensure this, all individuals hired to serve as FIs were required to read and sign a Data Collection Agreement (see Exhibit 3.1). Failure to comply with the provisions of this agreement could have resulted in termination from the 1999 NHSDA.

FI candidates who were unknown to the FS were interviewed by the FS. The FS fully explained the requirements and responsibilities of the NHSDA interviewer's job, described the project expectations, and defined the required time commitment. The FS then probed the candidate's job and interviewing history. At the conclusion of the interview, if the FS still considered the person a viable FI candidate, the FS conducted reference checks. If the reference checks were satisfactory, the FS then recommended the candidate for hire.

Because the paper-and-pencil (PAPI) supplemental sample for the 1999 survey oversampled Hispanics, FSs attempted to hire bilingual interviewers who spoke Spanish fluently in those sample areas with large populations of Hispanics. Before an FS hired a bilingual candidate, each applicant was screened by a bilingual staff member to assess the applicant's Spanish-language abilities. The assessment involved reading and speaking in Spanish. The bilingual candidate had to meet these assessment requirements satisfactorily before he/she could be hired and trained as an RTI-Certified bilingual interviewer.

Exhibit 3.2 displays a flow chart that presents all of the steps in the FI recruiting and hiring process.

During the entire data collection period, a total of 1997 FIs were hired and completed training. For those hired and sent to training, there were two sub-groups of FIs: those trained on the CAI instrument only, and those trained on the CAI and PAPI instruments, which constituted the majority of interviewers. The following are demographic characteristics of the trained staff:

• Of the 1997 FIs who were hired and completed training, 365 (18%) were Black or African-American and 149 (7%) identified themselves as "Other" (including Asian, American Indian, Pacific Islander, etc); 166 (8%) of those recruited were bilingual in Spanish.

• Of the 1497 trained in PAPI and CAI modes, 286 (19%) were Black and 126 (8%) were Other; 152 (10%) of those recruited for both PAPI and CAI were bilingual in Spanish.

• Of the remaining 500 FIs who were trained for CAI only, 79 (16%) were Black and 23 (5%) were Other; 14 (3%) of those recruited for CAI only were bilingual in Spanish.

Table 3.1 provides a distribution of interviewers by race and gender for the CAI instrument training only; Table 3.2 for the PAPI and CAI instrument training; and Table 3.3 for the total FIs trained. Table 3.4 provides a distribution of interviewers by bilingual skill and gender for the CAI only staff; Table 3.5 for the staff trained to work PAPI and CAI; and Table 3.6 for the total.

3.5 Problems Encountered

3.5.1 Original Plan vs. Actual Staffing Needs

The number of interviewers to be hired was based on a series of assumptions, including:

• the allocation of the sample across the FI Regions;



• the number of hours that an average FI would work each week, based on experience from previous NHSDA surveys;



• the estimated length of time to complete each screening;



• the estimated length of time to complete each interview; and



• the number of weeks that the interviewing staff would work in the quarter.

Based on these assumptions, RTI needed to hire and train approximately 1,200 interviewers for the beginning of data collection for Quarter 1. The number of staff who were hired and completed training was significantly less than the target. The reasons for not reaching the target include:

1. a robust economy which created competition for employees;

2. other field surveys competing for the same staff;

3. a limited pool of quality applicants who applied for the positions;

4. loss of staff who were hired but did not attend training for a variety of reasons; and

5. loss of staff during training because they were unable to complete the training or decided that the interviewing position was not suitable for them after they were exposed to the details of the position.

The effects of the reduced number of available staff were:

1. a decrease in overall production during Quarter 1;

2. the need for each FS to recruit new staff sooner than expected; and

3. the need to conduct an additional training session for new recruits.

As each quarter's sample was provided by the statisticians, the process to estimate the number of needed interviewers was repeated. The assumptions used to calculate the number of interviewers needed in each area were refined based on the most recent experience. For example, staffing plans were based upon an expected 15-20 hours per week per FI; certain areas experienced considerable difficulty obtaining this level of commitment so were able to increase the number of staff needed to ensure a sufficient amount of labor. Staff needed from quarter to quarter in an FS territory changed continuously, so FSs had to review staff assignments throughout the quarter and continually recruit and hire additional staff.

3.5.2 Staffing Shortfall

Throughout the year, the number of staff working continued to be less than the targeted number of interviewers needed. While some areas were close to the targeted number, other areas struggled. During Quarter 2, the concept of traveling interviewer teams was examined, then teams were created. Each RD region recruited a team of up to 10 interviewers from among their current staff and from newly-identified candidates with proven interviewing experience. These traveling field interviewers (TFIs) were hired at an out-of-pattern pay rate to compensate them for potentially not having as much work as regular interviewers. Each TFI was asked to commit to at least two 12-day trips each quarter. TFI teams were used to fill the unmet needs in areas with staffing shortfalls or where special needs arose (e.g., to cover long-term illnesses in the staff or to address problems created by very rich interview yields in some areas). In addition, several TFIs were certified bilingual interviewers and were assigned to areas where no bilingual interviewer was available.

3.5.3 Attrition

Attrition in the interviewing staff was higher than expected rising from 29.8% in 1998 to 38% in 1999. Proposed reasons for the increased attrition rate include:

1. staff taking jobs at higher rates of pay or with benefits;

2. staff determining that the interviewing job was not what they wanted to do;

3. staff determining that they were not capable of doing the job; and

4. staff not showing commitment to the survey by not working the number of hours they agreed to or by not working at the optimal times necessary for interviewing work.

The continuing attrition meant FSs had to recruit new staff more frequently than planned and continually juggle assignments to ensure that all of the assigned work was completed appropriately. There were significant costs associated with continuous recruiting efforts. These included not only the time of the FSs and the RTI office staff, but the costs of placing additional newspaper ads, preparing and shipping recruiting material, and travel to conduct interviews with candidates. Additional costs were also incurred when TFIs had to be sent to work in areas where no interviewer was available.  

Table 3.1

Distribution of 1999 NHSDA Interviewers ­ By Race and Gender

CAI

Race	Male	Percent Male	Female	Percent Female	Total	Percent of Total
Black	26	23.2%	53	13.7%	79	15.8%
White	81	72.3%	317	81.7%	398	79.6%
Other	5	4.5%	18	4.6%	23	4.6%
Total	112	100.0%	388	100.0%	500	100.0%

 

Table 3.2

Distribution of 1999 NHSDA Interviewers ­ By Race and Gender

PAPI

Race	Male	Percent Male	Female	Percent Female	Total	Percent of Total
Black	67	19.3%	219	19.1%	286	19.1%
White	245	70.4%	840	73.1%	1,085	72.5%
Other	36	10.3%	90	7.8%	126	8.4%
Total	348	100.0%	1,149	100.0%	1,497	100.0%

 

Table 3.3

Distribution of 1999 NHSDA Interviewers ­ By Race and Gender

Total

Race	Male	Percent Male	Female	Percent Female	Total	Percent of Total
Black	93	20.2%	272	17.7%	365	18.3%
White	326	70.9%	1,157	75.3%	1,483	74.3%
Other	41	8.9%	108	7.0%	149	7.5%
Total	460	100.0%	1,537	100.0%	1,997	100.0%

 

Table 3.4

Distribution of 1999 NHSDA Bilingual Interviewers by Gender

CAI

Language Ability	Male	Percent Male	Female	Percent Female	Total	Percent of Total
Bilingual	4	3.6%	10	2.6%	14	2.8%
Non-Bilingual	108	96.4%	378	97.4%	486	97.2%
Total	112	100.0%	388	100.0%	500	100.0%

 

Table 3.5

Distribution of 1999 NHSDA Bilingual Interviewers by Gender

PAPI

Language Ability	Male	Percent Male	Female	Percent Female	Total	Percent of Total
Bilingual	37	10.6%	115	10.0%	152	10.2%
Non-Bilingual	311	89.4%	1,034	90.0%	1,345	89.8%
Total	348	100.0%	1,149	100.0%	1,497	100.0%

 

Table 3.6

Distribution of 1999 NHSDA Bilingual Interviewers by Gender

Total

Language Ability	Male	Percent Male	Female	Percent Female	Total	Percent of Total
Bilingual	41	8.9%	125	8.1%	166	8.3%
Non-Bilingual	419	91.1%	1,412	91.9%	1,831	91.7%
Total	460	100.0%	1,537	100.0%	1,997	100.0%

 

Exhibit 3.1

Data Collection Agreement

Exhibit 3.2

Flow of FI Recruiting Activity

Exhibit 3.2(Continued)

Exhibit 3.2(Continued)

Exhibit 3.2(Continued)

Exhibit 3.2(Continued)

This page was last updated on June 03, 2008.

SAMHSA, an agency in the Department of Health and Human Services, is the Federal Government's lead agency for improving the quality and availability of substance abuse prevention, addiction treatment, and mental health services in the United States.