1999 National Household Survey on Drug Abuse Data Collection Report

6. DATA COLLECTION

This chapter presents the basic data collection procedures given to field staff working on the 1999 NHSDA. For further details or specific instructions, consult the 1999 NHSDA Field Interviewer Manual.

Interviewers were assigned specific sample dwelling units (SDUs) to contact, and the addresses or location descriptions were displayed on the Newton handheld computer. The sample was released in waves, with additional units made available as needed depending on progress during the initial weeks of data collection.

RTI made initial contact with residents of the specific SDUs by sending a lead letter signed by the project's National Field Director. The letter, which gave a brief explanation of the nature of the study and its methods, was mailed from RTI prior to the start of each quarter of data collection. All addresses were reviewed for completeness by a computer program, and SDUs lacking a complete mailing address were not sent a letter. To allow for these cases and other instances of delivery problems, each interviewer had extra copies to give to respondents during a personal visit. A copy of the letter, in both English and Spanish, was also included in the Showcard Booklet for reference.

In past years, interviewers received lead letters preprinted with the addresses of all SDUs. Each FI then reviewed the addresses for mailability, signed the letters, and mailed them throughout the first part of the quarter so that the letters arrived fairly close to the time the FI expected to be in the area. While the personal touch was a benefit, the mailing of lead letters was centralized to reduce FI burden and reduce costs.

Before knocking on the door of an SDU, the FI selected the appropriate case for that specific unit on the Newton. Each FI possessed a personalized letter of authorization printed on Department of Health and Human Services letterhead authorizing the FI by name to work on the study and approached the door with his/her RTI identification badge clearly visible. The FI also carried a variety of informational materials such as Question and Answer Brochures, NHSDA Highlights, and copies of newspaper articles about NHSDA.

When contacting the unit, the FI asked to speak with an adult resident (18 or older) of the unit who could serve as the screening respondent. The FI introduced himself/herself and the study. As scripted on the Newton screen, during the introduction the FI mentioned the lead letter and gave the screening respondent the Statement of Confidentiality. The Statement of Confidentiality, which was also included in the Showcard Booklet for reference, explained the purpose of the data collection effort, assured the respondent that all information gathered would be handled in the strictest confidence, and estimated the time required to complete the interview. The Statement also advised respondents that there were no known risks or benefits to their participation and that respondents were free to withdraw from the study at any time. Therefore, the Statement of Confidentiality provided all required aspects of Informed Consent for both the screening and interviewing portions of the study.

If no respondent was available or another situation was found at the unit so that screening could not be completed during the first visit, at least four callbacks were made to the unit so that each SDU was visited a minimum of five times in an effort to complete the screening. These contacts were made at different hours on different days of the week to increase the likelihood of completing the screening.

Screening was performed at each DU by obtaining information about the residents of the unit to determine whether or not any household member would be eligible for the NHSDA interview based on the ages of the DU members. The screening program guided the FIs through the process of asking age, gender, race/ethnicity, and military status for all persons aged 12 and older who lived at the unit for most of the calendar quarter, and the information was entered into the Newton.

Once the roster information was entered and verified, the FI started the within-dwelling unit selection algorithm on the Newton by tapping the "Make Selection" button. The Newton automatically determined, based on the composition of the household roster, whether or not anyone in the unit was selected for the interview. This tapping of one `button' on the Newton replaced the difficult paper screening process of determining selection, thereby eliminating errorsin the selection process. Project staff were quite pleased with the increased accuracy provided by the computerized selection.

The system allowed for the selection of none, one, or two members of a household for an interview. Dwelling units with 12- to 17-year-olds on the roster were more likely to have persons selected for an interview. It was possible that if two household members were chosen, they could be within the same age group.

In order to identify each selected individual, the Newton displayed the person's roster number (based on the order in which household members were listed), the age, gender, and either the relationship to the householder (for housing units) or a first name (for group quarters units). Also displayed was the mode of the interview: either a Computer Interview (CAI) or a Paper Interview (PAPI). Also listed for CAI interviews was a QuestID number, which was required to start the computerized interview on the laptop. For PAPI interviews, the FI had to instead enter the seven-digit QuestID number from the actual questionnaire booklet into the Newton. FIs transmitted all the completed screening data contained on the Newton to RTI each evening.

Once the selected individual(s) were identified during screening, the FI asked to complete the interview(s) at that time. If unavailable, the FI entered information about possible times for future contacts in the Newton Record of Calls. A minimum of four additional visits were made at different times of day/days of the week in an attempt to complete the interview.

For adults selected for the CAI interview, the FI used introductory scripts from the Showcard Booklet to introduce the study and the interview process. To meet the requirements of Informed Consent, the Statement of Confidentiality was provided as well. After receiving consent, the FI set up the computer equipment in a private location and began the interview.

For adults selected for a PAPI interview, the introductory scripts used by the FI were located on the inside page of the questionnaire booklet, and the Statement of Confidentiality was provided during the introduction. Once ready, the FI and the respondent determined a private location and began the interview.

If the selected individual was aged 12-17, the FI was responsible for obtaining verbal consent from a parent or guardian before contacting the youth. The only exceptions to this rule were in certain group quarters situations, like dormitories, where such consent was unobtainable, or if the youth was an emancipated minor. A separate paragraph for parents/guardians was included in both the CAI and PAPI introductory scripts. Once parental permission was granted, the FI approached the youth and introduced the study using the same adult script to obtain the youth's agreement to participate. Parents were then asked to leave the interview setting to ensure the confidentiality of the youth's responses. When ready, the FI and the youth began the interview.

The CAI interview began in the CAPI mode (computer-assisted personal interviewing), with the FI reading the questions from the computer screen and entering the respondent's replies into the computer. After completing the Reference Date Calendar, the FI explained to the respondent how to use the computer for the ACASI (audio computer-assisted self-interviewing) sections. Utilizing ACASI methodology for the sensitive drug usage questions enhanced privacy since the respondent listened to the pre-recorded questions through the headphones and entered the responses directly into the computer. Beginning with a practice session which introduced the various computer keys used during the interview, the respondent then proceeded through the interview. Four times during the ACASI portion of the interview, the respondent was instructed to ask the interviewer for a specific picture pill card designed to aid respondent recall. When the respondent was finished with the ACASI portion, the interviewer once again took charge of the computer, asking additional demographic questions as well as health care, insurance, and income questions. During both the beginning and ending CAPI portions, showcards were utilized to assist respondents in answering the questions.

The average CAI administration times overall and for the various sections of the CAI interview by respondent age (youth 12-17 or adult 18+) are given in Tables 6.1 through 6.40. Please note that the total number of interviews included varies between tables due to missing timing data, unresolved breakoff times or the exclusion of extreme values. For suspect timing data, the affected sections were not included in the analysis, though unaffected sections were included. Consider an example case: if timing for chewing tobacco was suspect, then the timing data for chewing tobacco, tobacco, and total time were excluded. Other non-suspect sections for that case, including other tobacco sections such as cigarettes, cigars, etc., were included in the analysis.

Interviewers conducting PAPI interviews used the tape-bound questionnaire booklet instead of the computer to guide the interview process. During the initial sections, the FI read the questions and recorded the answers in the booklet. After the Reference Date Calendar, the FI controlled the flow of the answer sheets that the respondent completed. All questions and answers were included in the questionnaire booklet in case the FI needed to answer respondent questions or read the items to assist a respondent with poor reading ability. As each answer sheet was completed, the respondent placed it in a large postage-paid envelope pre-addressed to RTI. Questions in the final demographic sections were then read by the FI, who recorded the responses in the questionnaire. When appropriate, showcards and picture pill cards were provided to aid respondent recall.

Table 6.41 gives the average administration times by respondent age group for the PAPI interview.

After the last interview question, the interview process for both CAI and PAPI involved several final steps. FIs had to:

• prepare the Verification Form and ask the respondent to complete the remaining items on the form;

• seal the completed Verification Form in a postage-paid envelope addressed to RTI;

• complete the FI Observation and Debriefing Questions;

• enter the final result code in the Newton;

• gather all interview materials; and

• thank the respondent.

All completed Reference Date Calendars were sent weekly to the field supervisor. Sealed Verification Form envelopes were mailed to RTI as soon as possible.

For CAI interviews, FIs also had to repack the computer equipment, and each night FIs transmitted interview data to RTI.

For PAPI interviews, the FI performed these additional tasks:

• scan edit of the completed questionnaire booklet;

• inserted the questionnaire booklet in the large envelope with the answer sheets;

• sealed the envelope and attached a strip of red non-removable "Confidential" tape; and

• invited the respondent to the nearest mailbox to ensure immediate mailing to RTI.

If the respondent did not accompany the FI to the nearest mailbox, the completed interview was mailed to RTI as soon as possible.

Project management on this massive study can be summed up in one word: communication. For instance:

• Interviewers throughout the country reported to their Field Supervisor twice each week to discuss production, problems encountered and possible resolutions, feedback on past work, plans for the next week, and any administrative issues.

• Field Supervisors each reported to their Regional Supervisor weekly, discussing production, costs, goals, staffing, and other administrative issues.

• Each Regional Director held a weekly meeting with his/her staff of Regional Supervisors to share project news and goals while addressing any problems within the region.

• All Regional Directors met each week with the National Field Director and the Project Director.

• All Directors and other key management staff met weekly with SAMHSA representatives.

Although the more formal meetings were held weekly, staff communicated almost constantly through the widespread use of e-mail. This management tool increased awareness of project issues by effectively passing information through the various management levels.

With the implementation of the Web-based project Case Management System, all management staff had access to a tremendous amount of information on the status of events in the field. Additional details on the CMS are provided in Section 8.2.

Another helpful management tool was the quarterly Performance Improvement Plan. At the end of each quarter of data collection, FSs developed specific plans in an effort to target particularly troublesome areas for improvement during the next quarter. Plans included the following information:

• A statement of the problem/situation to be addressed.

• A diagnosis of the problem in the past.

• Projected or desired outcomes.

• Specific efforts designed to accomplish these outcomes.

RSs assisted in the plan development and monitored the results of the plan's implementation.

At times during the data collection process, interviewers had difficulty gaining access to particular DUs. Interviewers with challenging circumstances were instructed to be observant, resourceful, and keep their supervisors informed of the situation. Additional suggestions taken from FS experience or from RTI's "Guide to Controlled Access Situations" were discussed. Talks with managers/owners generally centered on the importance of the study, SAMHSA and RTI's emphasis on confidentiality, and the right of the individuals to make a personal decision about participation. Supervisors sometimes contacted managers/owners directly to answer questions or concerns.

Due to prior efforts by staff who listed the DUs, many access problems were resolved readily. Listers recorded contact information and other steps followed to secure access so that interviewers could follow the same strategies or build on already-established relations. Supervisors at the listing stage used special reports on the CMS to monitor access situations; supervisors for screening and interviewing used the same reports and recorded additional information to update the reports.

For continuing problems, RTI had a system to generate individualized letters and packets of information about the project. When required, FIs and FSs provided basic information to RSs, who then requested the packets. Upon receiving the request, specialists at RTI prepared a cover letter and assembled materials to fit the situation. The packet was often sent via Federal Express to increase the importance placed on the contents and ensure timely delivery.

For persistent problem situations not resolved through FS/FI efforts or the letters/packets, "Please Call Us" letters were sent. Special care was taken that calls resulting from the letters were directed to the authorized RS or FS to set up an appointment so the FI could return and complete screening, or, in dire situations and with permission, screening information could be obtained by the FS or RS over the telephone.

Occasionally controlled access problems required assistance beyond the RS level so Regional Directors-and sometimes even the National Field Director-became involved.

More often than desired, potential respondents exercised their "right to refuse to participate." The following were in place to try to prevent refusal situations:

• The 1999 Field Interviewer Manual gave specific instructions to the FIs for introducing both themselves and the study. Additionally, an entire chapter discussed "Obtaining Participation" and listed the tools available to field staff along with tips for answering questions and overcoming objections.

• During FI training, two sections of the guide covered details for contacting dwelling units and how to deal with reluctant respondents and difficult situations. During exercises and mock interviews, trainees were able to practice answering questions and using letters and handouts to obtain cooperation.

• All aspects of the NHSDA were designed to exude professionalism and thus enhance the legitimacy of the project. All materials provided to the public were developed carefully. Interviewers were instructed to always behave professionally and courteously.

In refusal situations, staff followed these steps:

• Detailed notes describing the situation were recorded in a Refusal Report on the Newton. FIs classified the refusal according to one of seven categories.

• After transmission from the Newton to RTI, the category of refusal and any notes were then available to the supervisor on the Web-based CMS. The FI and FS could then discuss the situation, with the FS suggesting additional tactics if necessary.

• Once the refusal situation was discussed, a refusal conversion letter was sent (if appropriate). On the CMS, the FS selected a specific letter based on the stage of the case (screening or interviewing), the category of the reason for the refusal (too busy, confidentiality concerns, etc.) and, for interviewing, the person to be addressed (the actual respondent or the parent of a selected youth). The FS could also delete the request for the letter (in situations where a letter would not be helpful or could not be delivered) or release the letter for automatic production and mailing.

• The interviewer returned to the DU to try again with other tactics.

• Cases could be transferred to a different interviewer if necessary.

• Supervisors were available to reluctant respondents to discuss the importance of participation.

With the implementation of the 50 state design came problems not encountered in prior years. By selecting areas throughout the entire country, many different types of situations arose that had to be resolved. For example, the change to a national sample eliminated the need for oversampling black and Hispanic respondents. This change meant many more affluent neighborhoods were selected which increased the number of restricted and controlled access situations.

The large staff required by the size of the project meant many staff were either inexperienced or new to the NHSDA. For inexperienced management staff, often more time early on was spent learning than managing. Communication was vitally important, yet it was challenging to ensure that tips and suggestions were consistently conveyed to all staff.

Many of the FIs were new to interviewing and so had numerous basic interviewing skills to develop. Most interviewers were also new to the project and therefore had many, many details to learn. Some were inexperienced with computers and the vast majority had never used a touch screen like that on the Newton. All of these types and levels of inexperience required time to develop the skills necessary for competence. The rule of thumb from prior years was that it took one to two full quarters for an FI to learn and be skillful at his or her job. For staff conducting both CAI and PAPI interviews in 1999, the learning curve was even longer.

Even former NHSDA interviewers promoted to the FS role struggled in their new position. Others new to the project had content to master as well as supervisory skills to develop or enhance. Management tools-such as the project's Web-based CMS-were very useful, but required time for staff to become familiar and comfortable with using the wealth of information available to properly manage both cases and staff.

The constant turnover of interviewing staff meant there were not enough interviewers to adequately cover the assignments. Once replacement staff were in place, FSs underwent the learning curve process with these new FIs rather than being able to build on experience FIs had gained in the field. The continued attrition caused FSs to spend considerabletime dealing with staffing issues (recruiting, hiring, more intense supervision of new employee, etc.) and less time on appropriately managing the most difficult cases.

While refusals at the screening and interview level have historically been a problem for the NHSDA (as with all national-level household surveys), refusals were a more significant problem from the start of the 1999 NHSDA. Some factors contributing to the rise in refusals and corresponding decline in response rates were:

• Very favorable economic conditions meant members of selected households were employed at higher levels than in the past, at home less and less inclined to devote the necessary time to participate.

• The CAI component of the 1999 NHSDA did not oversample blacks and Hispanics-two groups that have cooperated at higher levels than others in previous years.

• The number of households with two persons selected for interview increased. Historically, response rates in households with two respondents are lower due to more frequent refusals by the second selected individual.

• Low unemployment rates caused a shortage of qualified and interested FI candidates. Those hired were often inexperienced as were many of the supervisors charged with managing the staff.

• Changing from paper/pencil to electronic screening prevented interviewers from manipulating the selection process so as to enhance the likelihood that a cooperative person was selected for the interview.

• The sophisticated CMS allowed for increased monitoring of questionable FI activities resulting in fewer fraudulent cases being submitted.

As is common in any large field data collection effort, staff encountered problems such as respondent availability, dwelling unit access (controlled or otherwise restricted), and high crime neighborhoods. Additionally, the use of escorts to increase interviewer comfort levels in unsafe areas had an impact on respondent reactions.

Using the Newton for electronic screening was a great use of new technology, but the Newton had its drawbacks:

• It was sensitive to a variety of weather conditions (and all types were encountered).

• As it became full of data, its response time slowed down and tried respondents' patience.

• The new technology created a confidence issue for staff who were unaccustomed to using computers.

• Concentrating on the device meant less eye contact with the respondent, which in turn made it tougher to establish good rapport.

Table 6.1

1999 NHSDA Time Analysis: Overall -
CAI Total Interview Time (Minutes)
with FI Debrief

Age Category	12-17	18+
Sample Size	22,377	35,555
Summary Statistics (in Minutes)
Mean	55.12	50.16
Variance	276.10	329.77
Standard Deviation	16.62	18.16
Quartiles
Maximum	201.28	224.82
Q3	64.85	59.03
Median	53.37	47.27
Q1	43.38	37.70
Minimum	9.70	9.92
Range	191.58	214.90
Mode	36.95	40.63
Percentiles
99%	103.50	108.87
95%	84.35	83.87
90%	76.58	73.08
10%	35.72	30.60
5%	31.78	26.80
1%	23.80	20.17
Extremes
5 Highest Values (Highest)	201.28	224.82
	190.02	206.65
	172.32	205.65
	166.78	204.08
	161.10	202.37
5 Lowest Values	12.08	10.48
	10.98	10.40
	10.53	10.22
	10.25	9.93
(Lowest)	9.70	9.92

Note: Time recording begins at screen STARTUP in the Introduction and stops recording after screen FIEXIT in the FI Debrief Module.

Table 6.2

1999 NHSDA Time Analysis: Overall -
CAI Total Interview Time (Minutes)
Without FI Debrief

Age Category	12-17	18+
Sample Size	22,783	36,193
Summary Statistics (in Minutes)
Mean	52.91	47.95
Variance	264.20	316.75
Standard Deviation	16.25	17.80
Quartiles
Maximum	198.60	219.35
Q3	62.42	56.73
Median	51.27	45.22
Q1	41.55	35.85
Minimum	8.92	8.63
Range	189.68	210.72
Mode	37.93	38.03
Percentiles
99%	99.70	105.17
95%	81.30	80.83
90%	73.85	70.35
10%	34.02	28.73
5%	29.90	24.85
1%	20.92	17.62
Extremes
5 Highest Values (Highest)	198.60	219.35
	185.25	201.50
	170.85	200.83
	158.72	200.23
	158.28	198.98
5 Lowest Values	10.17	9.27
	10.13	8.85
	9.78	8.77
	9.17	8.77
(Lowest)	8.92	8.63

Note: Time recording begins at screen STARTUP in the Introduction and stops recording after screen TOALLR3 in the Income Module.

Table 6.3

1999 NHSDA Time Analysis: Overall -
CAI Introduction (Minutes)

Age Category	12-17	18+
Sample Size	25,289	41,209
Summary Statistics (in Minutes)
Mean	3.79	3.96
Variance	5.60	6.73
Standard Deviation	2.37	2.59
Quartiles
Maximum	35.60	47.63
Q3	4.85	5.02
Median	3.37	3.45
Q1	2.15	2.23
Minimum	0.35	0.35
Range	35.25	47.28
Mode	2.08	3.75
Percentiles
99%	11.80	13.02
95%	7.85	8.38
90%	6.52	6.90
10%	1.37	1.47
5%	1.03	1.15
1%	0.72	0.80
Extremes
5 Highest Values (Highest)	35.60	47.63
	31.73	38.88
	30.37	38.77
	29.82	38.58
	28.88	36.80
5 Lowest Values	0.40	0.47
	0.37	0.47
	0.35	0.47
	0.35	0.47
(Lowest)	0.35	0.35

Note: Time recording begins at screen STARTUP in the Introduction and stops recording after screen CALENDAR in the Core Demographics Module.

Table 6.4

1999 NHSDA Time Analysis: Overall -
CAI Tutorial Section (Minutes)

Age Category	12-17	18+
Sample Size	25,335	41,305
Summary Statistics (in Minutes)
Mean	2.65	2.57
Variance	1.89	2.91
Standard Deviation	1.38	1.70
Quartiles
Maximum	27.65	29.83
Q3	3.45	3.28
Median	2.50	2.22
Q1	1.68	1.48
Minimum	0.03	0.05
Range	27.62	29.78
Mode	1.57	1.83
Percentiles
99%	6.60	8.18
95%	4.88	5.52
90%	4.27	4.50
10%	1.08	0.92
5%	0.77	0.63
1%	0.35	0.32
Extremes
5 Highest Values (Highest)	27.65	29.83
	24.43	28.78
	18.05	28.45
	17.75	27.55
	17.27	26.33
5 Lowest Values	0.13	0.10
	0.13	0.10
	0.08	0.10
	0.03	0.07
(Lowest)	0.03	0.05

Note: Time recording begins at screen INTRO1 in the Tutorial Module and stops recording after screen ANYQUES in the Tutorial.

Table 6.5

1999 NHSDA Time Analysis: Overall -
CAI Total ACASI (Minutes)

Age Category	12-17	18+
Sample Size	23,022	36,514
Summary Statistics (in Minutes)
Mean	38.28	33.21
Variance	190.60	222.50
Standard Deviation	13.81	14.92
Quartiles
Maximum	202.00	193.93
Q3	46.15	39.87
Median	36.48	30.37
Q1	28.57	23.12
Minimum	5.22	3.13
Range	196.78	190.80
Mode	26.12	20.85
Percentiles
99%	79.23	82.85
95%	62.47	61.55
90%	55.92	51.93
10%	22.72	17.97
5%	19.52	15.10
1%	13.12	9.85
Extremes
5 Highest Values (Highest)	202.00	193.93
	190.20	188.78
	164.17	183.07
	153.92	180.17
	152.85	168.03
5 Lowest Values	5.70	4.45
	5.70	4.27
	5.32	3.80
	5.23	3.58
(Lowest)	5.22	3.13

Note: Time recording begins at screen INTROACASI in the Tutorial Module and stops recording after screen ENDAUDIO in either the Youth Experience, Social Environment or Parenting Module.

1999 NHSDA Time Analysis: Overall -
CAI Total Core Section (Minutes)

Age Category	12-17	18+
Sample Size	25,218	41,105
Summary Statistics (in Minutes)
Mean	12.08	12.50
Variance	33.05	44.04
Standard Deviation	5.75	6.64
Quartiles
Maximum	70.87	88.93
Q3	15.32	15.45
Median	11.17	11.17
Q1	7.93	8.00
Minimum	0.53	0.72
Range	70.33	88.22
Mode	8.33	9.08
Percentiles
99%	29.47	34.47
95%	22.38	25.08
90%	19.62	20.92
10%	5.68	5.75
5%	4.50	4.55
1%	2.30	2.38
Extremes
5 Highest Values (Highest)	70.87	88.93
	56.38	88.92
	53.13	76.73
	51.47	76.10
	50.78	74.48
5 Lowest Values	0.73	0.80
	0.72	0.80
	0.65	0.80
	0.58	0.78
(Lowest)	0.53	0.72

Note: Time recording begins at screen LEADCIG in the Tobacco Module and stops recording after screen SV13 in the Sedative Module.

1999 NHSDA Time Analysis: Overall -
CAI Total Tobacco Sections (Minutes)

Age Category	12-17	18+
Sample Size	25,295	41,271
Summary Statistics (in Minutes)
Mean	2.99	3.18
Variance	2.96	4.45
Standard Deviation	1.72	2.11
Quartiles
Maximum	30.33	34.50
Q3	3.93	4.07
Median	2.70	2.77
Q1	1.73	1.77
Minimum	0.10	0.10
Range	30.23	34.40
Mode	1.45	2.50
Percentiles
99%	8.47	10.48
95%	5.97	6.87
90%	5.02	5.63
10%	1.18	1.07
5%	0.95	0.75
1%	0.50	0.35
Extremes
5 Highest Values (Highest)	30.33	34.50
	22.58	33.23
	21.93	31.38
	21.55	29.98
	19.85	29.02
5 Lowest Values	0.13	0.10
	0.13	0.10
	0.13	0.10
	0.12	0.10
(Lowest)	0.10	0.10

Note: Time recording begins at screen LEADCIG in the Tobacco Module and stops recording after screen CG43 in the Tobacco Module.

1999 NHSDA Time Analysis: Overall -
CAI Cigarette Section (Minutes)

Age Category	12-17	18+
Sample Size	25,345	41,331
Summary Statistics (in Minutes)
Mean	1.22	1.39
Variance	0.77	1.43
Standard Deviation	0.88	1.20
Quartiles
Maximum	18.98	27.58
Q3	1.53	1.87
Median	1.05	1.15
Q1	0.62	0.63
Minimum	0.03	0.02
Range	18.95	27.57
Mode	0.50	0.17
Percentiles
99%	4.03	5.38
95%	2.80	3.33
90%	2.27	2.68
10%	0.43	0.22
5%	0.35	0.15
1%	0.23	0.07
Extremes
5 Highest Values (Highest)	18.98	27.58
	16.92	26.08
	16.90	24.78
	16.12	24.45
	15.60	24.03
5 Lowest Values	0.05	0.02
	0.05	0.02
	0.05	0.02
	0.03	0.02
(Lowest)	0.03	0.02

Note: Time recording begins at screen LEADCIG in the Tobacco Module and stops recording after screen CG16a in the Tobacco Module.

1999 NHSDA Time Analysis: Overall -
CAI Chewing Tobacco Section (Minutes)

Age Category	12-17	18+
Sample Size	25,348	41,338
Summary Statistics (in Minutes)
Mean	0.46	0.52
Variance	0.29	0.47
Standard Deviation	0.54	0.68
Quartiles
Maximum	29.82	28.40
Q3	0.47	0.58
Median	0.33	0.30
Q1	0.22	0.18
Minimum	0.02	0.02
Range	29.80	28.38
Mode	0.43	0.18
Percentiles
99%	2.72	3.20
95%	1.37	1.70
90%	0.78	1.15
10%	0.15	0.12
5%	0.12	0.10
1%	0.05	0.05
Extremes
5 Highest Values (Highest)	29.82	28.40
	14.88	27.37
	10.52	15.37
	8.07	12.83
	7.97	11.95
5 Lowest Values	0.02	0.02
	0.02	0.02
	0.02	0.02
	0.02	0.02
(Lowest)	0.02	0.02

Note: Time recording begins at screen CG17 in the Tobacco Module and stops recording after screen CG24 in the Tobacco Module.

1999 NHSDA Time Analysis: Overall -
CAI Cigar Section (Minutes)

Age Category	12-17	18+
Sample Size	25,312	41,298
Summary Statistics (in Minutes)
Mean	0.29	0.37
Variance	0.18	0.22
Standard Deviation	0.42	0.47
Quartiles
Maximum	9.72	15.23
Q3	0.23	0.52
Median	0.15	0.17
Q1	0.08