The Impact of School Climate and School Identification on Academic AchievementMultilevel Modeling with Student and Teacher Data

A teacher's experience of their school — whether they feel respected, purposeful, and part of something — shows up in their students' test scores. Not because the teacher delivers some attitude directly. Not because of anything said in the classroom. Just the fact of the teacher's perception, measured separately, predicts how well students read and calculate. Let that sit for a beat. Then say: Maxwell, Reynolds, Lee, Subasic, and Bromhead went looking for the mechanism. School climate is not a vague feeling. It's a concrete, multi-dimensional psychosocial construct — the quality and character of school life, including norms, values, expectations, and the pattern of relationships that shape how learning happens. Researchers have settled on three core dimensions that appear repeatedly across the literature: academic emphasis, meaning how much the school is genuinely driven by a pursuit of learning; interpersonal relationships, meaning the quality of connections among students and between students and staff; and shared norms and values, the common understandings of purpose and acceptable behavior. Safety and disorder — bullying and aggression — tend to appear within those relational and normative dimensions or as separate subscales. The measurement problem is real. Dozens of different instruments exist, each with different subscales and emphases, which makes it hard to compare findings across studies. And perspective matters: students, teachers, principals, and parents can perceive the same school very differently. Most research has handled this by asking one group — almost always students. That's where the field has been stuck. Relying solely on student self-report creates two compounding problems. First, you get a partial picture — students see a lot, but not everything that shapes a school's character. Second, when you measure both your predictor and your outcome from the same rater, you risk conflating the rater's general disposition with actual school effects. Methodologists in the field have argued for years that multi-informant measurement is best practice. Several prior studies — including work by Johnson and Stevens, and Brand and colleagues — had linked teacher perceptions of school climate to student outcomes. But almost no study had simultaneously integrated staff and student perceptions into a single analysis and asked whether each carries independent signal. That gap is what Maxwell and colleagues set out to close. The theoretical engine for their study comes from the social identity approach, built from social identity theory by Henri Tajfel and John Turner and extended by Turner and colleagues through self-categorization theory. The core idea is a sharp psychological distinction between personal identity and social identity. When a person identifies with a group, that group's values and goals become personally meaningful. The group's norms become their norms. Applied to schools: when students come to see their school as a meaningful social group — as "us" — they internalize its academic goals and exert the kind of effort that changes outcomes. School climate is the input. School identification is the gear that converts climate into behavior. A school with strong relationships, clear shared values, and genuine academic emphasis is a school students can identify with. And identification, the authors predicted, is where the mechanism lives. Now for the study itself. Maxwell and colleagues collected data from two thousand two hundred fifty-seven students and seven hundred sixty staff across seventeen public secondary schools in the Australian Capital Territory — eighty-nine percent of the district's public high schools. Students were in Grade 7 and Grade 9, with a mean age of thirteen point three years. Academic achievement was measured using NAPLAN, Australia's national literacy and numeracy assessment, standardized on a scale from zero to one thousand. The sample means were five hundred sixty-five for numeracy, five hundred seventy-three for reading, and five hundred twenty-five for writing. Students completed the SCASIM-St, a thirty-eight item climate survey covering academic emphasis, staff-student relations, student-student relations, and shared values, plus a six-item school identification scale. Staff completed a mirrored instrument, the SCASIM-Sf. Both showed strong internal reliability — school identification alphas above zero point ninety-four for both groups. The analyses used two-level structural equation modeling in MPlus, explicitly separating variance at the individual student level from variance at the school level. Student-level controls included grade, gender, parental education, and language spoken at home. School-level controls included ICSEA — the school's socioeconomic index, ranging from approximately nine hundred seventy-two to one thousand one hundred seventy-eight across the sample — plus school size and survey response rate. Because there were only seventeen schools, staff climate scores were averaged to the school level and then disaggregated onto individual records, allowing the model to test whether staff perceptions explained student-level variance beyond what students themselves reported. Here's what the numbers showed. For students: perceptions of school climate predicted stronger school identification, with a path coefficient of about one point ten across models — consistent and statistically significant. School identification then predicted higher achievement. For numeracy, the identification coefficient was ten point zero three; for writing, ten point eight nine. Both significant at the five percent level. The indirect effects — climate to achievement running through identification — were ten point nine for numeracy and eleven point eight three for writing, both significant. Reading showed a similar pattern but at a marginal significance level. In plain terms: students who perceived a more positive school climate identified more strongly with their school, and that identification translated into measurably higher scores. Now for the staff finding, which is the result that raises eyebrows. Staff-perceived school climate significantly predicted student achievement even after accounting for students' own perceptions. In the full model, staff perceptions predicted writing with a coefficient of twenty-one point twenty-one, reading at sixteen point eighty, and numeracy at seven point fifty-seven — all significant at the five percent level. That full set. Across numeracy, reading, and writing, while student responses were already in the model. The combined models explained around forty-two percent of total variance in numeracy scores, about thirty-nine percent in reading and writing, with student-level climate and identification accounting for roughly six to nine percent of within-school variance depending on the outcome. Staff perceptions were not redundant with student perceptions. They carried independent signal. Then there's the null result, which is just as informative. Staff school identification — how strongly teachers identified with their school as a group — did not predict student achievement and did not moderate the relationship between staff-perceived climate and outcomes. The asymmetry is clean and somewhat puzzling: student identification mediates the student-climate effect, but staff identification does nothing detectable on the staff side. The authors flag two likely reasons. First, seventeen schools is fewer than most methodologists recommend for stable multilevel moderation tests. Second, staff identification showed very little variability — a mean of five point eighty-eight on a seven-point scale with a standard deviation of only one point twelve. When nearly everyone scores near the top of a scale, the statistical leverage to detect effects disappears. So the null on staff identification is real but should be held lightly until tested in a larger, more variable sample. What does this mean for schools? The practical upshot is that school climate is a lever for student outcomes — and that the lever has two handles. For students, the mechanism runs through identity and belonging. Improving how students experience their school's relationships, academic emphasis, and shared values doesn't just make school feel better; it makes students feel like they belong to something, and that sense of belonging changes how they engage with learning. For staff, the channel is different — or at least unresolved. Staff perceptions of climate independently predict how students perform, but we don't yet know why. It's not staff identification. The mechanism on the staff side remains open. Maxwell and colleagues briefly point to the ASPIRe model as a practical framework — an intervention approach designed to strengthen organizational identification by improving the quality and salience of shared group membership. Applied to schools, that would mean explicitly designing conditions that make the school a meaningful social group for students: coherent values, visible academic purpose, genuine relationships. The study's cross-sectional design means we can't yet say that improving climate causes identification to rise and scores to follow. But the mediation pattern is consistent with that causal story, and the authors make a clear call for longitudinal and experimental tests. Seventeen schools is a starting point, not a ceiling. The next studies need larger, more diverse samples, genuine variation in staff identification, and designs that can track whether deliberate climate interventions shift identification first and achievement second. The psychological mechanism is plausible, the evidence is coherent, and the question of what teachers experience when they walk into work turns out to matter more than most school research has bothered to ask. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.