III. Data Collection
A. Time Frame of the Survey
The Bureau of the Census collected the 1994-95 TFS data during the 1994-95 school year. Table III-1 summarizes the specific data collection activities and the time frame in which each occurred.
Table III-1. Data collection time schedule
------------------------------------------------------------------------------- Activity Month of Activity ------------------------------------------------------------------------------- Advance letters mailed to LEAs and state administrators Aug-94 Teacher Status Forms (TFS-1) and letters mailed to sampl Sep-94 Reminder postcards mailed to sample schools Sep-94 Telephone followup of Teacher Status Forms not returned October-November 1994 Initial mailing of leaver/stayer questionnaires (TFS-2 a Jan-95 Second mailing of leaver/stayer questionnaires (TFS-2 an Feb-95 Telephone followup of mail questionnaire nonrespondents March-May 1995 --------------------------------------------------------------------------------
The U.S. Bureau of the Census collected TFS data in two phases:
Phase 1 - In September 1994, the Census Bureau mailed teacher status forms (TFS-1) to schools that had provided lists of teachers for 1993-94 SASS. On this form, the school principal (or other knowledgeable school staff member) was asked to report the current occupational status of each teacher who participated in the 1993-94 SASS, by indicating whether he/she was still at the school in a teaching or nonteaching capacity, or had left the school to teach elsewhere or for a non-teaching occupation. One week after the TFS-1 mailout, reminder postcards were mailed to the sample schools. In October and early November, Census interviewers telephoned schools that had not returned the TFS-1 to obtain the requested information.
If the school staff indicated a sample teacher had moved, the Census Bureau also tried to obtain address correction information for the sample teacher's home address through the post office.
Phase 2 - In January 1995, the TFS questionnaires were mailed to selected teachers and former teachers. The Questionnaire for Former Teachers (TFS-2) was sent to sample persons reported by school administrators as having left the teaching profession. The Questionnaire for Current Teachers (TFS-3) was sent to sample persons who were reported as still teaching at the elementary or secondary level. When home addresses were available, the questionnaires were mailed there. Otherwise, they were mailed to the sample teacher's 1993-94 school.
In February, the Census Bureau mailed a second questionnaire to each sample person who had not returned the first questionnaire. Also, for those who returned the first form and indicated that it did not apply to them (because their status was incorrectly reported by their 1993-94 school), the appropriate questionnaire was mailed to them. For example, if a sample person who was teaching in another school received the questionnaire for former teachers (TFS-2), he/she was instructed to return the questionnaire; the Census Bureau sent the correct questionnaire (TFS-3 for current teachers) to the respondent during the second mailout.
In late March 1995, Census interviewers began calling sample persons who had not returned a mail questionnaire. In addition to these nonresponse followup cases, some "nonmailable" cases (cases with incomplete addresses) were assigned for telephone followup. If the interviewers were unable to contact a sample teacher through a contact person (two contact persons had been listed by the sample teacher on the SASS form as knowing how or where to get in touch with him or her) or through directory assistance, they called the sample person's 1993-94 school to obtain information about the person's current address or employer. Telephone followup of nonrespondents ended in May 1995.
During the telephone followup, the interviewers used form TFS-4, Telephone Questionnaire for the Teacher Followup Survey 1994-95, to collect the data. This form was a combination of the TFS-2 and the TFS-3 questionnaires; i.e., data for a current or a former teacher could be recorded on this single form.
IV. Edit Procedures
Questionnaires returned by individual respondents and those completed by Census interviewers during telephone followup were sent to the Census Bureau processing center in Jeffersonville, Indiana. Upon receipt, clerks assigned codes to each questionnaire to indicate its status e.g., complete interview, refusal, deceased. Then they performed a general clerical edit that included reviewing all entries for legibility and making corrections, e.g., changing "one" to "1" and rounding fractions to whole numbers. For the TFS-2 (Questionnaire for Former Teachers), the clerks also assigned industry and occupation codes/9 to the respondent's current job (questionnaire item 3). For TFS-3 (current teachers) respondents teaching in a different state, they assigned a state FIPS code to the respondent's new state (questionnaire item 13).
After the clerical edit, the questionnaires were batched by type and interview status (i.e., interviews, noninterviews, out-of-scope for survey) for data keying. Data entry personnel were instructed to correct all errors identified during the keying operation and to refer problem cases to their supervisor. To assure the quality of the data, all keying was independently verified at the 100 percent level.
B. Preliminary ISR Classification
After keying data, the next step in processing was to make a preliminary determination of each case's interview status (ISR) i.e., whether it was an interview, a noninterview, or out-of-scope for the survey. In general, those cases with "out-of-scope" check-in codes (assigned by clerks when questionnaires were received at the processing center) were classified as out-of-scope (ISR=3) for the preliminary ISR.
Otherwise, cases with data entries were classified as interviews (ISR=1) and those with no data were classified as noninterviews (ISR=2).
After the assignment of the preliminary ISR code, the file was divided into two files: (1) former teachers (leavers) and (2) current teachers (stayers and movers). Then these files were submitted to a computer edit that consisted of a range check, a consistency edit, and a blanking edit. Only records classified as interviews in the preliminary ISR were edited.
The range check deleted entries that were outside the range of acceptable values.
The consistency edit identified inconsistent entries within each record and, whenever possible, corrected them; if they could not be corrected, the entries were deleted.
Example: If the respondent answered "Yes" in Item 9 to indicate that he/she had earned a college degree during the past 12 months but recorded a date in Item 10 (month and year of degree) that was later than June 1995, the "Yes" response in Item 9 was changed to "No" in the consistency edit and the entry in Item 10 was deleted.
The blanking edit deleted extraneous entries and assigned the "not answered" (.N) code to items that should have been answered but were not.
D. Final Interview Status Edit
After the range check, consistency edit, and blanking edit were completed, the records were edited to determine whether the case was eligible for inclusion in the survey and, if so, whether sufficient data had been collected for the case to be classified as an interview. An interview status recode (ISR) value was then assigned to each case as a result of the edit:
Item 1 - Main occupational status
Item 4 - Full-time/part-time employee status
Item 5 - Earnings at new job
Item 8 - Main activity for next year
Item 18 - Whether respondent plans to return to teaching
Item 23a - Main reason for leaving teaching profession
Item 5a - Full-time/part-time teaching status
Item 7a - Main teaching assignment field
Item a - Whether respondent is certified in main teaching assignment field
Item 9a - Whether respondent is certified in other teaching assignment field
Item 10 - Grade levels of students taught by respondent
Item 11 - Type of community where school is located
Item 12 - Whether respondent has changed schools
Item 28 - Main activity for next year
Item 51b(1) - Teaching salary
Item 51c - Total earned income
-------------------------------------------------------------------------------- Teachers in sample Interviewed Not interview Out-of-scope -------------------------------------------------------------------------------- Public 4,528 463 34 Current teachers 2,798 281 2 Former teachers 1,730 182 32 Private 1,751 330 17 Current teachers 1,175 232 0 Former teachers 576 98 17 BIA 44 6 0 --------------------------------------------------------------------------------
NOTE: Noninterviews and out-of-scope teachers are not included in the TFS analytic or public-use datafiles.
SOURCE: U.S. Department of Education, National Center for Education Statistics, "Teacher Followup Survey, 1994-95," (TFS-2 Former Teachers Questionnaire, and TFS-3, Current Teacher Questionnaire) unpublished tabulations.
--------------------------------------------------------------------------------------------------------------------------------------
Table V-1 summarizes the weighted and unweighted response rates for the TFS (shown in percentages).
---------------------------------------------------------------
Interviews
---------------------------------------------------------------
Sampled Teachers Unweighted Weighted
---------------------------------------------------------------
Public 90.7 92.3
Current teachers 90.9 92.5
Former teachers 90.5 89.2
Private 84.1 87.2
Current teachers 83.5 87.2
Former teachers 85.5 87.6
BIA 88.0 99.5
Current teachers 95.5 99.9
Former teachers 82.1 88.9
Total 88.8 91.6
Current teachers 88.6 91.8
Former teachers 89.1 88.8
---------------------------------------------------------------
Table V-2. Survey response rates for 1993-94 SASS Teacher List, 1993-94 Schools and Staffing Teacher Survey, and 1994-95 Teacher Followup Survey, weighted
-------------------------------------------------------------------------------------------------
Sector
-----------------------------------
Public Private
Current Former Current Former
Components Teachers\10 Teachers Teachers Teachers
--------------------------------------------------------------------------------------------------
SASS Teacher List response rate\11 95.0 91.0
SASS Teacher Survey response rate\12 88.2\13 80.2\14
Teacher Followup Survey response rate\15 92.5 89.2 87.2 87.6
--------------------------------------------------------------------------------------------------
A cumulative overall response rate is the product of the survey response rates shown in Table V-2; (SASS Teacher List response rate) (SASS Teacher Survey response rate) (TFS Teacher response rate). The cumulative overall response rates by sector and teacher status for the 1994-95 TFS are:
Public current teachers: (.95)(.882)(.925)(100)=80.0.
Public former teachers: (.95)(.882)(.892)(100)=74.7.
Private current teachers: (.91)(.832)(.872)(100)=66.0.
Private former teachers: (.91)(.832)(.876)(100)=66.3.
Table V-3 is a brief summary of the unweighted item response rates for the 1994-95 TFS questionnaires. A response rate for an item is defined as the number of records with valid responses to that item divided by the number of eligible respondents for the item.
|
|
Table V-3. Summary of Unweighted Item Response Rates
---------------------------------------------------------------------------
Former teacher Current teachers
---------------------------------------------------------------------------
Range of item response rates 78-100% 52-99%
Percentage of items with a response
rate of 90% or more 92.60% 90.40%
Percentage of items with a response
rate less than 80% 1.20% 4.10%
Items/16 with a response rate less than 7 9b, 9c, 48i, 51a(1),
80% 51a(2), 51a(3),
51b(3), 51b(4), 57b
--------------------------------------------------------------------------
FOOTNOTES:
[9] The lists of industry and occupation codes used for TFS-2 item 3 are provided in Appendix D of this publication.
[10] Includes stayers and movers.
[11] Percent of schools providing teacher lists for the 1993-94 SASS sample, unweighted.
[12] Percent of eligible sample teachers responding to the 1993-94 SASS Teacher Survey, unweighted.
[13] This rate does not include the 5 percent of the public schools which did not provide teacher lists.
[14] This rate does not include the 9 percent of the private schools that did not provide teacher lists.
[15] Percent of eligible sample teachers responding to the 1994-95 Teacher Followup Survey, weighted.
[16] The questionnaire wording for these items can be found in Appendices B and C of this publication.
Technical Notes