The big idea
Before you observe a client or run a single assessment, the client already has a history on paper: school reports, medical charts, past evaluations and earlier behavior plans. A record review is the systematic reading of those documents at the start of a case to learn what has already been found, tried and ruled out.
A good review saves time and protects the client. It can reveal a medical condition that explains the behavior, a plan that failed for a clear reason, or a skill the client already uses in one setting. It also tells you which questions your own assessment still has to answer.
Breaking it down
Getting the records
Records are confidential. You need written permission from the client or their legally authorized representative before requesting or exchanging them (Standard 2.11, Obtaining Informed Consent; see E.4). Ask for the documents most likely to matter, store them securely and read them with a purpose.
Educational records
School documents show how the person performs where many demands are placed on them.
- Individualized education programs (IEPs) or family service plans, with current goals and services
- Previous evaluations: psychoeducational, speech-language, occupational therapy
- Earlier functional behavior assessments and behavior plans, with any data on how they worked
- Attendance, discipline referrals, report cards and changes of placement
Medical records
Medical information comes first in importance, because some behavior is partly driven by biological variables that no behavior plan will fix. Look for:
- Diagnoses, especially conditions linked to pain or discomfort (ear infections, constipation, reflux, dental problems, menstrual pain, allergies)
- Current medications, recent changes and known side effects
- Seizure history, sleep problems, and hearing and vision results
- Hospitalizations, injuries and any limits on activity or diet
Standard 2.12 (Considering Medical Needs) asks you to make sure medical needs are assessed and addressed when there is any reasonable chance they influence the behavior, to document referrals and to follow up.
Historical records
History puts the current behavior in context.
- Developmental milestones and when concerns first appeared
- Past services, how long they lasted and what changed
- Major life events: moves, losses, new siblings, changes in caregivers or residence
- Legal documents such as guardianship orders, which tell you who can consent
What to do with what you find
A record review is an indirect assessment: the information was gathered by other people, at other times, often for other purposes. Treat it as a source of hypotheses, not conclusions. Records can be out of date, incomplete or written in nonbehavioral terms (“gets frustrated”). Note possible setting events (poor sleep, illness, a new medication), list what has been tried, and carry these leads into interviews and direct observation (F.5).
Seeing it in practice
Darnell, 9, has started hitting his head during class. His family consents to a record review.
| Record | Finding | What it suggests |
|---|---|---|
| Pediatric chart | Two ear infections in three months; last visit was before the head hitting began | Possible pain-related behavior: refer for a medical check before assuming a social function |
| IEP | Goals for asking for help; weekly speech services | Few ways to ask for help or a break |
| Earlier behavior plan | Time-out for head hitting, with no data recorded | No evidence it worked; function was never assessed |
| Parent intake form | Head hitting is rare on weekends | Ask what differs between school days and weekends |
The review does not tell the analyst why Darnell hits his head. It tells them where to look first, and it makes a medical referral the first step.
In a second case, Mei, 34, lives in a group home and recently began refusing meals. Her medication list shows a new prescription that started the week the refusals began. With consent, the analyst shares this with the house manager and Mei’s physician before designing any feeding intervention.
Common mix-ups
- A diagnosis is not a function. A record may say a client has an anxiety disorder. That tells you nothing about what maintains a particular behavior. You still need F.5 and F.6.
- “Tried and failed” needs a second look. A plan may have failed because it was never implemented well, never matched the function or never had data. Check before ruling an approach out.
- Records do not replace seeing the person. They are where assessment starts, not where it ends.
- Medical first does not mean medical only. Ruling out a medical cause rarely means waiting for behavioral help. Often both run at the same time.
Exam tip: When a scenario mentions a sudden change in behavior, a new medication, signs of pain or a behavior that appeared after an illness, the best first step is usually to consult or refer to a medical professional, not to begin a functional analysis.
Quick recap
- Get written consent before requesting or sharing records.
- Educational, medical and historical records answer different questions.
- Medical variables come first because they can drive behavior and change what treatment is needed.
- Records are indirect assessment: use them to form hypotheses, then test those hypotheses directly.
- Document what you reviewed, what you found and any referrals you made.
Further reading
- Cooper, J. O., Heron, T. E., & Heward, W. L. (2020). Applied behavior analysis (3rd ed.), chapters 3 and 27.
Key terms
- Record review
- Systematic reading of a client's existing documents, such as school, medical and historical records, at the start of a case to learn what has already been found, tried and ruled out.
- Indirect assessment
- Assessment that gathers information about a behavior from sources other than direct observation of it, such as interviews, rating scales, questionnaires and records. In a functional behavior assessment it is used to form hypotheses, not to confirm function.
- Setting event
- A condition or event, often earlier in time or ongoing, that changes how likely a behavior is in a later situation, for example by altering how effective a consequence is. Many behavior analysts now analyze these effects as motivating operations.
- Informed consent
- Permission given by a person with the legal right to consent, after they understand what they are agreeing to: the purpose, procedures, risks and benefits, any limits on confidentiality, and their right to decline or withdraw at any time. It is required before services begin, before substantial changes, and before information is shared or used publicly.
Check yourself
5 questions on F.1. Choose an answer to see whether you're right and why.
Show the answer
Answer: C. A sudden change in behavior that lines up with a medication change is a reasonable sign that a biological variable may be involved, so a medical review comes first. A functional analysis is tempting, but it would test social consequences while a possible medical cause goes unexamined.
Show the answer
Answer: B. At 19, Aiyana consents for herself unless a court has appointed a guardian, so a guardianship order answers the question directly. An IEP or evaluation may mention who attends meetings, but it does not establish legal authority to consent.
Show the answer
Answer: B. A plan can fail because it was implemented poorly, never matched the function, or was never measured. Without data, the label 'unsuccessful' cannot rule the approach out. Concluding that time-out can never work goes beyond what the record shows.
Show the answer
Answer: B. Records were written by other people, at other times and often for other purposes, so they describe behavior secondhand. They still require consent to obtain. Testing hypotheses by changing antecedents describes an experimental analysis, not a record review.
Show the answer
Answer: B. Physical discomfort can make problem behavior more likely and change how well consequences work, so it is a lead to share with the physician and test, not a conclusion. It does not prove any function, and medical and behavioral work can often run at the same time.
Adds a step to your study record. Saved in this browser only.