Report pages 146–155: 143 claims on pp. 146–155 (fact 100, statement 19, evidence 1, conjecture 1, definition 22).
In this appendix, we describe our methods for administering a survey to a nationally representative sample of K–12 public school principals via RAND’s American School Leader Panel (ASLP) and for analyzing the results of that survey.
As our methodology for collecting and analyzing survey data remains relatively consistent across ASLP surveys, the description of our methods below is updated from a previous publication analyzing results from another ASLP survey focused on school safety (Jackson et al., 2023).
The ASLP is a panel of K–12 public school principals.
It was designed to produce national estimates from the perspective of public school principals, as well as subgroup estimates by respondent and school characteristics.
Principals are recruited into the ASLP using probability-based sampling methods.
The sampling frame for the ASLP is a list of U.S. public school principals purchased from Market Data Retrieval (MDR), a market-research firm specializing in education.
The ASLP currently includes about 8,000 principals.
Principals who have joined the panel have agreed to participate in surveys several times per year and receive incentives for completing surveys.
After being recruited into the ASLP, panelists are then invited to participate in specific ASLP surveys using probability-based sampling methods.
Survey items were designed in-house by RAND researchers and shaped by prior work on school safety and security.
Although the survey sponsor provided feedback on survey items, RAND researchers maintained editorial control and were responsible for institutional review board review and approval of the final survey.
The survey was intentionally designed to gather information from both principals in schools with and without BTAM teams. More specifically, the survey was designed to allow respondents to progress through different pathways depending on their specific school context, as shown in Figure B.1.
Respondents were sorted onto one of the pathways depending on their responses to some initial screener items.
The main (longer) pathway was designed to collect detailed information from respondents in schools with BTAM teams. It was anticipated that this pathway would take respondents 20 minutes to complete.
Respondents in schools that did not have BTAM teams (or had only an external team) progressed through a shorter survey, designed to take at the most 15 minutes to complete. These participants received more scenarios about discipline incidents at their schools.
1 | Introduction (screening questions) | Module 2 (Yes / No / Don’t know)
2 | Scenarios (n = 3 / n = 5 / n = 5) | Module 3; Module 8 or Module 7; Module 7 or END
3 | Inside BTAM | Module 4 if involved in initial implementation (Yes), else Module 5 (No)
4 | BTAM Implementation | Module 5
5 | BTAM Operations | Module 6
6 | BTAM Sustainability | END if “Outside BTAM” = no; Module 7 if “Outside BTAM” = yes
7 | Outside BTAM Teams | END
In the second module of the survey, respondents received scenarios that described a behavioral incident that might occur at school.
Respondents were asked to read the scenario and then were asked the following questions about how their school would respond to such an incident:
Which person or group is most likely to make the initial decision about how to respond to this student’s behavior?
Below is a list containing a wide variety of options schools can use to respond to concerning student behaviors, depending on their seriousness. Most incidents would involve only a subset of these options. What action(s) would this person or group at your school likely take or recommend in response to this student’s behavior?
Would the BTAM team(s) serving your school be made aware of this student’s behavior?
The follow-up questions were tailored depending on the school’s situation. For example, schools that reported not having a BTAM team did not receive the third question asking whether their BTAM team would be made aware of the incident.
The personnel (e.g., nurses, social workers, school-based law enforcement) that respondents saw who might respond to an incident reflected only those personnel that respondents said they had at their school.
Table B.1 lists the 16 scenarios used in the survey. All scenarios were randomly assigned.
As shown in Figure B.1, the respondents on the BTAM-intensive survey pathway received three scenarios, while those on the shorter survey pathway received five scenarios.
The scenarios were designed to include a mix of things, including severity of the incident (low, medium, or high) and student populations affected (e.g., special education student, general education student).
Low-severity scenarios described incidents in which BTAM teams likely should not be involved (e.g., low-level discipline incidents, such as classroom misbehavior).
High-severity scenarios described incidents in which BTAM teams definitely should be involved, such as students making direct threats to themselves or others.
Medium-severity scenarios were designed to describe borderline cases, in which schools might have more discretion over whether BTAM teams can or should be involved.
All respondents received one low-severity scenario and one high-severity scenario. Most respondents then received one medium-severity scenario, although some respondents received three.
Respondents saw their scenarios (and the associated follow-up questions) in a random order.
The scenarios were scaled to the grade level of students in the schools that principals reported serving.
Principals were first categorized into school grade levels based on their reports of the grade levels of students their schools served in that school year.
Principals were categorized as primarily serving middle grades if their school contained grades 7 or 8.
Principals were categorized as serving primarily elementary grades if their school contained any grades from kindergarten through 6.
Principals were categorized as serving primarily high school grades if their school contained any grades from 9 through 12.
If principals were in a school that served both elementary and high school grades, they were randomly assigned to be elementary or high.
Principals in schools serving mostly elementary grades were assigned the elementary scenarios, while principals in schools serving mostly middle or high school grades were assigned the secondary scenarios.
This was to ensure that respondents received scenarios that felt more plausible for the type of incidents their schools might actually encounter in their jobs.
A | Low | Elementary | A student is repeatedly disruptive in class, wandering the classroom, interrupting the lesson by talking to other students, and sometimes shouting loudly to get attention. The student does not respond to multiple teacher requests to stop their behavior.
B | Low | Elementary | A student regularly engages in disruptive behavior during class, but their teachers do not consider their behavior threatening. Last school year, the family declined staff suggestions that the student might benefit from supports for neurodiverse learners. This school year, the student’s behavior has escalated, and the student is involved in a physical altercation with another student that does not result in injuries.
C | Medium | Elementary | A student gets into a pushing altercation with another student on the bus. After the altercation is broken up by the bus driver, the student says they were responding to having been tripped while they were walking down the aisle. During the discussion with the student, the bus driver believes they smell alcohol on the student’s breath. The parents of the other student involved in the altercation call the school the next day concerned for their child’s safety on the bus.
D | Medium | Elementary | A student has an intense verbal confrontation with another student at school, shouting, “I am going to kill you for that.” The student has no history of fighting or disruptive behavior but has had academic problems. A teacher who spoke with the student does not believe the threat was serious but does note that the student started crying during their conversation and would not share why they were so upset.
E | Medium | Elementary | In a drawing assignment that asked students to visualize different ways to solve problems, a student draws a graphic picture of someone shooting at a group of people, including details like bleeding and other physical injuries. The teacher mentions previous unsuccessful attempts to contact the student’s family about their academic progress, and that the student had recently appeared withdrawn and sad in class.
F | Medium | Elementary | A student has a history of emotional outbursts and non-compliance with directions, including behaviors involving misuse of and damage to school property (e.g., pushing books off shelves, throwing objects). After a teacher attempts to redirect the behavior and takes their materials away, the student shouts, “You will be sorry tomorrow you did that.” The school established there were firearms in the home when responding to past behavioral incidents involving same student.
G | Medium | Elementary | A peer reports that a fellow student has been repeatedly writing them disturbing notes. The student continues to send the notes after having been asked to stop and does not seem concerned about the effects they have been having on the other student. The notes have become more concerning over time, including descriptions of inappropriate contact, but do not include any explicit threats. The student is known to come from an unstable home environment, and the school became aware last year that the student was making repeated comments to their peers about a recent school shooting.
H | High | Elementary | A student has a history of getting into fights with other students at school and being generally disruptive in the classroom. Earlier this year, the student complained about being singled out and bullied by many other members of the class. The student was later caught bringing a knife to school, explaining it was “for protection,” and an investigation determined that there were firearms in the family home. After a fight on the playground with a classmate, the student yells that “he’d be sorry” and warned another classmate that “they shouldn’t come to school tomorrow.”
I | Low | Secondary | A student is repeatedly disruptive in their biology class, talking with other students or playing on their phone. Though in a course normally taken by students one grade below theirs, the student still appears to be struggling with the material. The teacher has made attempts to correct or redirect the behavior but has been unsuccessful and reached a point where frustration is affecting the entire class.
J | Medium | Secondary | A student regularly misses classes and is repeatedly late to class. The student talks back to teachers in an aggressive tone when they are confronted about their tardiness and absences. After hearing breaking glass in a school bathroom, a hall monitor checks in to find a broken beer bottle on the bathroom floor; the student had just been seen exiting the bathroom.
K | Medium | Secondary | During P.E., two students begin pushing each other during a soccer activity. When the P.E. teacher approaches and tells them to stop, both back off and apologize to varying degrees. The teacher notices and is concerned that one of the students still appears angry and agitated after the fight.
L | Medium | Secondary | A teacher observes a student in the hallway between classes showing off a pocketknife to a group of other students. When questioned, the student replies they only wanted to show their new knife to their friends. No other students present at the time report that the student made threats or engaged in other concerning behavior. The student voluntary relinquishes the knife to the teacher when asked.
M | Medium | Secondary | A student regularly distracts peers with side conversations during class and rarely completes assignments despite having a 504 plan for additional academic support. After sustained disruptive behavior, the teacher moves the student to another seat in the classroom, which causes the student to complain they were treated unfairly. The teacher is asked to come to the office to discuss the situation with the student and school staff, at which point the student begins yelling profanities at the teacher. Last year, the student needed to be restrained after becoming violent with a teacher during a school field trip.
N | Medium | Secondary | A student notices that another student in class is especially interested in the details of the Columbine High School shooting and the personal history of the students involved. The student repeatedly brings up the attackers by name during discussions with peers and has been searching for information about the incident on their school laptop. The teacher has also noticed that the student has become increasingly withdrawn in class.
O | Medium | Secondary | A classmate reports that another student at school is connected to a social media account that has posted threats targeting the school. A review of the account finds posts with pictures of weapons, discussions of grievances justifying violence, and indicators the author may be suicidal. The posts do not include specific details of planned violence but do carry an increasingly urgent and desperate tone. The student has also become increasingly isolated at school, and their grades have dropped. No information other than the classmate’s word connects the student to the account.
P | High | Secondary | A fight breaks out in the hallway during passing time, and student witnesses reported hearing one of the aggressors shout, “see you tomorrow for another round,” and the bullied student replying, “Fine. This ends tomorrow.” The incident involves a student who has a history of being bullied and was caught brandishing a knife (which they said they needed “to try to get all of them to stop”) last school year. An investigation determined that there were unsecured firearms in the family home.
Our principal survey was administered online in English between February 5, 2025, and February 24, 2025.
Invited principals received multiple reminder emails to participate.
Participants received an incentive upon completion of the survey, which was scaled to correspond to the length of the survey ($1 per minute).
That is, those who received the longer survey received a $20 incentive, while those who received the shorter version received $15.
To minimize the likelihood that respondents would change their survey answers to receive the higher incentive, respondents were not told which pathway they were on or which survey items triggered each pathway.
Before going into the field, researchers identified two main goals for analysis. The first goal was to produce estimates that generalize to the national population of U.S. public school principals and the schools in which they serve.
The second goal was to facilitate comparisons in how schools operationalized their BTAM programs across three state policy contexts: (1) states that had early legislation (as of 2019), (2) states that encouraged BTAM use, and (3) other states.
Early legislation states (light blue) | WA, OR, NV, TX, MS, GA, FL, IL, MI, OH, PA, NY, CT, NH, KY, VA, MD, RI
States that encouraged BTAM use (purple) | MT, ID, ND, UT, CO, NE, KS, OK, NM, MO, AR, TN, IN, WV, NC, SC, VT, MA, NJ, DE
Other states (orange) | CA, AZ, WY, SD, MN, IA, WI, LA, AL, AK, HI, ME, DC
SOURCE: Features data compiled in 2019 by Child Trends and NASBE (2021).
Our sample was designed to support these two main goals while considering several other factors.
The first consideration was researchers needed to ensure sufficient sample to unpack differences in schools’ BTAM programs with some uncertainty about how many schools nationally had BTAM programs as of January 2025.
Using data collected from the Department of Education’s School Pulse Panel in April 2024, we estimated that about eight in ten sampled schools would have a BTAM team (NCES, 2025).
A second consideration was ensuring sufficient sample size to estimate means and proportions across the three state policy contexts of interest without substantially increasing the survey design effect.
A final consideration was the availability of panelists across state policy contexts within the ASLP to be sampled. There were about 6,000 active panelists in the ASLP at the time of sampling.
The final sample was chosen to target 1,650 completed surveys distributed across the three state policy contexts.
Sample (no n: a percentage change, not a share of a sample) · a percentage that is not a share of a sample (e.g. a percentage-point change or a weighted estimate of the population)
The initial goal was to sample 50 percent from the early legislation states (as it was anticipated that BTAM programs would be most common in these states), 25 percent from each of the other states, and 25 percent from the encouraged BTAM use only states.
Sample (no n: a percentage change, not a share of a sample) · a percentage that is not a share of a sample (e.g. a percentage-point change or a weighted estimate of the population)
As shown in Table B.2, assuming a 30 percent response rate (which is in line with historical ASLP surveys), this would require sampling 1,375 principals in both the other states and encouraged BTAM use only states and 2,750 principals in early legislation states.
Sample (no n: a percentage change, not a share of a sample) · a percentage that is not a share of a sample (e.g. a percentage-point change or a weighted estimate of the population)
There were not enough panelists in the other states in the ASLP at the time of sampling to meet the desired sample. Therefore, we sampled some additional principals from the encouraged BTAM use only states to maintain the 50 percent/50 percent distribution with principals from early legislation states.
RAND researchers used probability-based sampling to invite 5,500 ASLP members to complete the survey using the ASLP as the sampling frame.
Of the 5,500 members sampled, 1,665 principals fully completed the survey, and 181 principals partially completed the survey.
Meanwhile, 38 sampled principals were determined ineligible for the survey.
Sample (no n: a percentage change, not a share of a sample) · a percentage that is not a share of a sample (e.g. a percentage-point change or a weighted estimate of the population)
Excluding these 38 principals and the cases in which respondents did not complete a significant portion of the survey, a total of 1,746 cases received a weight. This corresponds to a 32.0 percent completion rate using the American Association for Public Opinion Research’s Response Rate 6 definition (American Association for Public Opinion Research, 2023).
Sample (n = 1320 of 1650) · Targeted number of completed surveys (design target)
We anticipated that about 80 percent of the 1,650 completes would receive the full survey and be asked detailed questions about their BTAM program (n = 1,320) and 20 percent would receive the shorter survey (n = 330).
However, we underestimated the number of principals who would indicate that their school had another “school-based team that makes decisions regarding serious student behavioral concerns that is not referred to as a behavioral assessment and management (BTAM) team” and receive the longer survey pathway.
About 250 schools told us that they have a school-based team that makes decisions regarding serious student behavioral concerns that is not referred to as a BTAM team.
This means that virtually all schools received the longer survey, and only about 50 schools received the shorter pathway intended to accommodate schools that were not doing BTAM.
Sample (n = 423 of 1307) · Table B.2 final sample, other states
Other states | 412 | 1,375 | 1,307 | 1,307 | 423 | 32.4
Sample (n = 481 of 1443) · Table B.2 final sample, encouraged BTAM use only states
Encouraged BTAM use only | 412 | 1,375 | 1,794 | 1,443 | 481 | 33.3
Sample (n = 842 of 2750) · Table B.2 final sample, early legislation states
Early legislation states | 826 | 2,750 | 3,219 | 2,750 | 842 | 30.6
Sample (n = 1746 of 5500) · Table B.2 final sample, all states
All | 1,650 | 5,500 | 6,320 | 5,500 | 1,746 | 31.7
Principals’ survey responses have been weighted to be representative of the national population of public school principals at the time the survey was administered.
To provide a weighted sample similar to the population of K–12 public school principals in the United States, we created sampling weights.
The weighting process accounts for the probability of selection into the survey from the panel and the likelihood an invited principal completes the survey (the response weight), and these likelihoods are combined and calibrated to reproduce the population distribution of K–12 educators.
More specifically, the final analysis weights used in our analyses are the product of the following three interim weights:
calibrated weight of the sampling frame (a calibration weight that assigns a weight for each panelist based on individual- and school-level characteristics so that the sum of the weights along the calibration factors closely match the national characteristics of the national population of public school principals based on the NCES’s National Teacher and Principal Survey)
sample selection weight (the inverse of the probability of selection into the survey sample using the panel as the frame)
survey response weight (the inverse of modeled probability of a principal completing the survey. The response adjustment is important to eliminate observed sources of response bias).
The products of these weights were subsequently recalibrated and trimmed if necessary.
As shown in Table B.3, both before and after weighting, survey respondents look similar to the population of U.S. public school principals on a variety of person-level and school-level characteristics.
We report samplewide and subgroup-specific means and proportions of variables of interest, weighted using the survey weights we describe in the previous section.
Because principals’ perceptions and experiences can vary depending on their school context, we examined differences in educators’ responses by school characteristics (in addition to their state policy context).
More specifically, we examined differences by school grade level, school and district enrollment size, locale, poverty status, and student racial/ethnic composition.
We obtained data on school and district demographics by linking survey data files to the Common Core of Data (CCD) and the Education Demographic and Geographic Estimates (EDGE) School Neighborhood Poverty Estimates, both issued by NCES.
school grade level (elementary, middle, and high)
school locale (urban, suburban, and rural)
school student racial/ethnic composition (schools in which more than half of students were Black, Hispanic, Asian, Pacific Islander, American Indian/Alaska Native, or of two or more races were categorized as having majority students of color, with the remaining schools categorized as having majority White students.)
Our definition of locale aligns with the four-category definition of locale used by NCES, with the exception that we collapsed those located in towns into the rural category.
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
Urban | 25.8 [23.8, 27.9] | 27.7 [25.4, 30.1] | 27.7
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
Suburban | 32.3 [30.1, 34.5] | 30.9 [28.5, 33.2] | 31.0
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
Town/rural | 41.9 [39.6, 44.2] | 41.4 [38.9, 43.9] | 41.3
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
Low (401 or more IPR) | 25.1 [23.1, 27.1] | 23.4 [21.3, 25.5] | 23.3
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
Medium (201 to 400 IPR) | 59.7 [57.4, 62.0] | 58.4 [55.9, 61.0] | 58.4
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
High (0 to 200 IPR) | 15.2 [13.5, 16.9] | 18.2 [16.1, 20.3] | 18.3
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
0%–50% | 55.8 [53.5, 58.1] | 51.0 [48.4, 53.6] | 50.8
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
50%–100% | 44.2 [41.9, 46.5] | 49.0 [46.4, 51.6] | 49.2
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
Elementary | 55.6 [53.3, 57.9] | 59.1 [56.6, 61.6] | 59.1
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
Middle | 22.3 [20.4, 24.3] | 17.9 [16.1, 19.8] | 17.8
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
High | 22.1 [20.1, 24.0] | 22.9 [20.7, 25.1] | 23.0
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
Small | 50.3 [48.0, 52.7] | 51.4 [48.8, 54.0] | 51.4
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
Large | 49.7 [47.3, 52.0] | 48.6 [46.0, 51.2] | 48.6
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
White | 79.4 [77.5, 81.3] | 78.3 [76.2, 80.5] | 78.2
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
Person of color | 20.6 [18.7, 22.5] | 21.7 [19.5, 23.8] | 21.8
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
Female | 49.7 [47.4, 52.1] | 55.5 [52.9, 58.0] | 55.5
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
Male | 50.3 [47.9, 52.6] | 44.5 [42.0, 47.1] | 44.5
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
Master’s or less | 47.8 [45.5, 50.2] | 63.4 [61.1, 65.7] | 63.6
Sample (N = 1746) · All principals who completed the survey and received a weight (weighted to national K-12 public school principals/schools)
Higher than master’s | 52.2 [49.8, 54.5] | 36.6 [34.3, 38.9] | 36.4
Population data for educator-level characteristics come from the National Teacher and Principal Survey, and school-level characteristics are from the CCD and the EDGE School Neighborhood Poverty Estimates.
Numbers in brackets represent confidence intervals.
school neighborhood poverty (low, middle, or high poverty based on the income-to-poverty ratio [IPR] for the neighborhood surrounding the school.
IPR is the percentage of family income that is above or below the federal poverty threshold set for the family’s size and structure and ranges from 0 to 999.
Schools with an IPR less than or equal to 200 were categorized as high poverty.
Schools with an IPR of greater than or equal to 401 were categorized as low poverty.
Remaining schools were categorized as middle poverty.
school enrollment size (schools with fewer than 450 students were categorized as small, whereas the remainder were categorized as large.)
district enrollment size (districts serving fewer than 10,000 students were categorized as small, whereas districts serving 10,000 students or more were categorized as large.)
It is important to keep in mind that each principal who took our survey is in a school that belongs to multiple of these categories—for example, a principal in a small, rural elementary school in a small district that serves majority White students in a relatively lower-income neighborhood in a state that does not address BTAM in state policy.
Thus, any patterns observed across multiple school contexts might be driven by principals in the schools that share multiple of these characteristics.
Sample (no n: a percentage change, not a share of a sample) · a percentage that is not a share of a sample (e.g. a percentage-point change or a weighted estimate of the population)
In this report, we describe only those differences among principal subgroups that are statistically significant at the 5 percent level, unless otherwise noted.
Because of the exploratory nature of this study, we did not apply multiple hypothesis test corrections.
Find the Perfect Training Class For You
