4.4 Ethical Assessment and Avoiding Bias

4.4 Ethical Assessment and Avoiding Bias

PECT lens: Objective 0002.5 tests ethical practices in assessment and strategies for avoiding bias. On this exam, ethics is not a soft extra. It is whether you use assessment to help children learn or to rank, expose, and punish. Pennsylvania teachers are also bound by 22 Pa. Code Chapter 235 (Code of Professional Practice and Conduct for Educators), the NAEYC Code of Ethical Conduct (especially responsibilities to children and families), and FERPA plus Pennsylvania student-records rules on confidentiality.

A technically perfect running record that is posted on the classroom door with a red name is still an ethics failure. A "rigorous" vocabulary quiz that only children who watch a particular American cartoon can pass is still a biased measure. PECT will give you both kinds of stem.

What ethical assessment looks like in PreK–4

Ethical practice starts before the first item:

1. Assess for a legitimate educational purpose tied to a standard or to identifying need — not to fill a bulletin board or to sort children into public status groups.
2. Use tools that can actually show the skill for this age and this child (Sections 4.1–4.2).
3. Protect dignity during administration: no sarcasm, no reading scores aloud, no using a quiz as revenge for behavior.
4. Interpret in context (language, culture, disability, health, opportunity to learn).
5. Act in the child's interest — reteach, support, communicate with families — rather than using scores to exclude from recess, field trips, or the next grade as a first move.
6. Share results only with people who have an educational need to know, in a setting that is private.

NAEYC's ethical stance is that assessment should benefit children and improve instruction. A PreK documentation panel that helps the teacher plan the next investigation is ethical. A preschool "readiness test" used to bar a child from kindergarten without multiple indicators or a team process is not. Do not use assessment to shame. Color-coded name charts, "test police" jokes, and posting ranked scores are climate harms and confidentiality harms at once.

Confidentiality of results is a legal duty, not a courtesy. FERPA protects personally identifiable education records, including grades, standardized scores, and much classroom performance data when they are maintained as records. Families have rights to inspect and to control disclosure. Colleagues who need the information to teach the child (the reading specialist who will group with you; the IEP team) may share it in professional spaces. The teacher next door who is gossiping in the hallway, the parent volunteer at the copy machine, and the class across the hall looking at a wall of percentiles do not. Pennsylvania professional conduct rules reinforce that educators do not disclose student information except for a professional purpose. On items, the right move is a private conference, a sealed folder, a password-protected gradebook, or a team meeting — never a public ranking.

Avoiding bias: four lenses PECT actually uses

Bias here means the assessment measures something other than the intended standard in a way that systematically disadvantages a group or a child. Avoiding bias is not "being nice." It is validity.

Bias also hides in opportunity to learn: judging children on a standard they were never taught, or comparing a late-enrolling student to peers who had three extra months of the core program, then calling the gap a child deficit. Ethical interpretation asks whether the child had a fair chance to learn the content.

Construct assessments with this checklist:

• Does every item require the target skill, or extra cultural/linguistic baggage?
• Could a child with a disability show the skill if the format changed but the standard stayed?
• Would a child without extra money, travel, or a quiet home still have access?
• Did I try the task myself for hidden idioms and assumptions?
• Will I interpret a dual language learner's errors as language until proven otherwise, not as a quick disability label? (Overrepresentation is 0003.8; the ethical seed is here.)

Scenario: linguistic bias in Grade 2 science

Ms. Brooks writes a quiz item: "The plant was under the weather, so it drooped." An English learner who can explain wilting from the investigation still misses the item. The item measured an idiom, not STEELS-aligned science. Ethical repair: rewrite in plain academic language, allow a diagram response, and do not put "science below basic" on a report based on that item.

Scenario: socioeconomic bias in kindergarten

A "draw your bedroom" prompt is used to assess spatial words. Children in doubled-up housing or shelters either freeze or are scored as lacking spatial language. The teacher switches to "draw a place you sleep or rest" and also assesses on/under/beside with classroom blocks. The standard stayed; the biased assumption left.

Teaching to the test versus teaching the standard

Teaching the standard means children spend time doing the actual work the Pennsylvania standard names: reading connected text, explaining mathematical thinking, investigating, writing for a purpose. Aligned practice, including some familiarity with how a CRT item looks in Grade 3–4, is legitimate when it does not replace the curriculum.

Teaching to the test means narrowing instruction to released item formats, test-taking tricks, and only the content likely to appear on a high-stakes form — at the expense of untested but required standards, play, science investigations, and the arts. It is unethical because children lose opportunity to learn the full standard, and because scores then reflect coaching on the instrument rather than the learning SAS is built to produce. On PECT, the correct option is usually the teacher who uses PSSA or benchmark results to find standards that need better instruction — not the teacher who cancels science for eight weeks of bubble practice.

Authentic, formative, and benchmark tools (4.1) all still must be aligned. Alignment is not the same as item-drill.

Accommodations are not cheating

An accommodation changes how the child accesses the assessment (extra time, a scribe, a separate setting, text-to-speech where allowed, a calculation device specified in a plan) so the score can reflect the target skill. A modification changes what is expected. Withholding an IEP or 504 accommodation "so we know what they can really do" is unethical and illegal: you then measured the disability or the barrier, not the standard. Giving an accommodation to a child who does not need it to inflate a class average is also unethical. The ethical rule is simple: provide what the plan requires; do not treat access as a reward or a cheat.

Do not confuse this with 0002.6 (legally acceptable modifications and ELPS-designed content assessments), which is the next chapter. For 0002.5, remember the ethics sentence: accommodations level access; they are not cheating; using assessment to shame is never "motivating."

Scenario: the hallway chart

Mr. Hale posts a named green/yellow/red benchmark chart by the door "to motivate." Children in red are teased at recess; a parent volunteer photographs the chart. Ethical violations: confidentiality (FERPA-relevant disclosure), shame, and using assessment as public ranking. Repair: keep the data in a private gradebook, form small groups quietly, and conference with families privately. Motivation comes from instruction the child can succeed at, not from a traffic light with a name on it.

Scenario: the colleague who says "scribe is cheating"

A Grade 4 student with an IEP scribe accommodation dictates a social studies short answer. A colleague wants the score thrown out. The ethical response is that the standard was content understanding, not handwriting stamina. The scribe was required access. Throwing out the score would punish the child for using a legal accommodation.

Exam traps for 0002.5

• Public scoreboards as "formative." They are shame and a confidentiality breach.
• Bias only happens on the PSSA. Classroom quizzes carry linguistic, cultural, disability, and socioeconomic bias too.
• Teaching to the test = being rigorous. Rigorous means teaching the standard, not the bubble.
• Accommodations = cheating or optional extras for nice days.
• Talking about a child's scores with anyone who asks. Need to know, private setting, family rights.
• One biased task used as a high-stakes label.

If an option protects the child's dignity, privacy, access, and opportunity to learn the actual standard, it is usually the 0002.5 answer. If it uses assessment to expose, exclude, or fake purity of scores, it is the trap.