Questions to Ask When Analyzing Student Data
For teachers, grade-level teams and school leaders sitting down with assessment results, grades and attendance in a data meeting. The questions follow the order the conversation needs to take: what the numbers show and what they cannot, which students and groups are behind, how the material was taught, what the errors have in common, what to reteach, and how you will check that it worked.
Want questions from the whole vault instead? Try the random question generator.
The questions
Each question, and why to ask it
The data
What does this assessment actually measure, and how many questions test each skill?
Why ask it
Put the test itself on the table before the spreadsheet. A useful answer names the standards or skills the items covered and how many items sat on each. If all anyone can say is 'reading' or 'math', the team is about to draw conclusions about skills that two questions cannot support.
What do we notice in the results before we try to explain anything?
Why ask it
Give this a few minutes of plain observation: 'eleven students missed item 6', not 'they did not study'. When someone says 'because', write it down as a guess for later and go back to what the page shows. Teams that skip this step tend to leave with the explanation they walked in with.
What did students do well, and which skills look secure?
Why ask it
Start a list of what is secure before the list of what is not. It tells you what to leave out of the reteaching and which lessons from this unit to keep as they are. If the only strength anyone can name is 'they tried hard', go back to the item report and find the questions nearly everyone got right.
What did we expect to see, and where do the results surprise us?
Why ask it
Have each person jot a prediction before the results go up. A surprise marks the place where what you believed about a class and what happened part ways, so start the discussion there. If nobody is surprised by anything, ask whether this assessment told you something the gradebook had not.
How many students does each percentage on this page stand for?
Why ask it
In a group of six, the move from 50 percent proficient to 67 percent is one student. Ask for the count beside every percentage, and be slow to call anything a trend in a group small enough to list from memory.
Which students have no score here, and what do we know about why?
Why ask it
Absent on test day, enrolled late, tested in another room, exempted: each is a different story, and the students who miss assessments are often the ones the meeting most needs to talk about. An average without them flatters the class. List them by name and decide who follows up.
Were the conditions the same for every class and every student?
Why ask it
Compare the time allowed, paper or screen, whether directions were read aloud, and the day it was given. If one room had forty minutes and another had two class periods, the gap between them is about conditions, and the fix is a shared set of directions next time. Check too that students with accommodations received them.
Do the test results agree with grades and with what we see in class?
Why ask it
When a student has high grades and a low score, or the reverse, look at what the grade is made of. Completion, behavior and extra credit can carry a grade a long way from what a student knows. Treat the disagreement as the finding and do not average it away.
What are we comparing these results to, and is that comparison fair?
Why ask it
Last year's class was different children, the room next door may have a different mix, and a different test is a different ruler. The steadiest comparison is usually the same students against the standard or against their own earlier work. Settle the yardstick before anyone says 'up' or 'down'.
Is this a real change, or the kind of bounce we see from one test to the next?
Why ask it
Scores wobble with a hard passage, a fire drill or a bad night's sleep. A dip of a few points for one class on one test rarely deserves a new plan, while the same dip on three checks in a row does. Look at how far earlier results moved from one check to the next before reading meaning into this one.
What can these numbers not tell us?
Why ask it
A score says a student missed an item. It does not say why, how hard they tried, or whether they could have done it with one prompt. Name the two or three things the team will have to learn another way, by reading the work or asking the student, so the spreadsheet is not asked to answer them.
Students
Which students are well below where they need to be, by name?
Why ask it
Nobody can teach '32 percent', so get from the percentage to a list as fast as you can. A good meeting ends this question with names on paper and an adult beside each one. If the list is longer than the help available, say so and rank it instead of pretending everyone will be reached.
Which students are just short of the mark, and what exactly is each one missing?
Why ask it
These students often need one skill, and the item report can usually tell you which. Be honest about why you are asking. Picking the students nearest a cut score because they lift the pass rate, while those further down wait, is a choice a team should make out loud or not make.
Which students already have this, and what will they be doing while we reteach?
Why ask it
This is the group a data meeting forgets. The worrying answer is 'they can help the others' every time. Look for a task that goes deeper into the same standard, not a longer set of the problems they just got right.
When we break the results out by student group, where are the gaps?
Why ask it
Use whichever groups your school reports on, such as students learning English or students with a support plan, and ask how each is defined there, since the labels differ from place to place. Check the counts before the percentages. Read a gap as a question about what those students were offered, never as a description of them.
Which students are growing even though they are still below the target?
Why ask it
A student who went from far behind to a little behind made real progress, and a proficiency line hides it. Name them, find out what was done for them, and tell them. Then look for the opposite: a student above the line whose scores have not moved since the fall.
Which students have dropped since the last check?
Why ask it
A fall is more urgent than a steady low score, because something changed. Before planning extra instruction, ask the adults who see that student every day what else moved: attendance, schedule, seat, friends. One low result after several strong ones calls for a conversation with the student first.
For students with a failing grade, is it missing work or low marks on the work they turned in?
Why ask it
Missing work is a matter of time, organization or attendance, and a short list of the assignments that count most will move the grade. Low marks on finished work mean the material was not learned, and chasing assignments will not fix that. Ask how your school's grading rules treat late and missing work before promising a student anything.
Who shows up on more than one list: low scores, falling grades and absences?
Why ask it
Lay the three lists side by side. A student on all of them needs one plan and one adult, not three interventions from three people who have not spoken. If attendance slipped first, tutoring on the missed content will not hold until the absences are dealt with.
How much of the instruction were the lowest-scoring students actually present for?
Why ask it
Count the days they attended during the unit, then subtract time out of the room for services, testing or discipline. A student who was gone for a large part of it has not failed to learn the material. They need it taught for the first time, and someone should ask what keeps pulling them out.
For each student on this short list, what has worked for them before?
Why ask it
Go around the table for each name, and ask anyone who taught the student in an earlier year. A useful answer is something a teacher could do tomorrow, such as a seat near the front, the problem read aloud, or ten minutes with one particular adult. Share only what the people present need, and ask what your school's rules are on discussing student information.
Which students have come up at every meeting this year without their results changing?
Why ask it
The answer tests the meetings, not the students. A name that has appeared four times beside the same plan means the plan should change: another approach, another adult, or a referral through whatever process your school has for extra support.
Teaching
How was this taught in each classroom, and for how many days?
Why ask it
Ask each teacher for a one-minute account: materials, time spent, how students practiced. Different results between rooms only mean something once you know what differed in the teaching. If people start defending themselves, the question was heard as a ranking, so say plainly that it is not one.
Is the weak spot in one classroom, across the grade, or in more than one grade?
Why ask it
One room is a conversation between colleagues about how that unit went. A whole grade suggests the materials or the pacing, and several grades suggest a skill that no single year has been given the job of teaching. A school leader looking across teams should settle this before asking anything of one teacher.
Where did one class do clearly better, and what did that teacher do differently?
Why ask it
Borrowing what worked next door is the cheapest improvement on offer. It depends on the teacher being specific ('I had them sketch the problem before touching the numbers') and on the others being able to ask without it stinging. Rule out class mix and testing conditions first so you are not copying luck.
Did the assessment ask for the skill in the form students had practiced it?
Why ask it
Students who practiced with multiple choice and then met a written response may know more than the score suggests. If format is the gap, a few rounds of practice in the new format will do more than teaching the content again. If the standard itself calls for the harder format, then that is what was not taught.
Which items covered material we had not reached, or had rushed through?
Why ask it
Mark these before anyone reads the item report. Low scores on them are a fact about pacing, and they belong in a separate pile from things that were taught and not learned. When that pile is large, the meeting should be about the calendar.
How much did students practice this on their own before they were tested?
Why ask it
A lesson where the teacher models and the class answers together can look fine until each student is alone with the page. Ask what the independent work was and whether anyone looked at it. If the first time students tried it alone was the assessment, you have found something to change before the reteach.
What did we see during the unit that told us this was coming?
Why ask it
Think of exit tickets, homework and the questions students asked. If the signs were there, find out what stopped anyone acting on them, which is often pressure to keep pace. If there were none, build a midpoint check into the unit before it is taught again.
Could any of these questions be flawed, with confusing wording, two defensible answers or a mistake in the key?
Why ask it
An item that nearly everyone missed, strongest students included, deserves a second read before it earns a reteach. Have someone who did not write it work it cold. Drop a bad item from the analysis and note it, so the same question does not mislead next year's team.
What earlier skills did this unit assume students already had?
Why ask it
The miss is often a grade or two back: a fractions unit that fails on multiplication facts, an essay that fails at the sentence. A good answer names the earlier skill and a quick way to check it. Repeating the grade-level lesson will not help a student who is missing the step underneath it.
Errors
Which items or standards did the most students miss?
Why ask it
Sort the results by item instead of by student. A handful of items usually accounts for most of the lost points. See whether they share a standard, a question type, or a place near the end of the test, where running out of time looks exactly like not knowing.
On the most-missed question, which wrong answer did students choose, and what thinking leads there?
Why ask it
When most students land on the same wrong answer, they probably share a misconception, and you can teach against it directly. Wrong answers spread evenly look more like guessing, which suggests the idea did not land at all. Work the item yourself the way a student would have to in order to reach that answer.
What do the errors have in common across students?
Why ask it
Put six or eight pieces of real student work on the table and sort them into piles by kind of mistake. Give each pile a few words, such as 'adds the denominators' or 'retells instead of arguing'. A pile with a name is a lesson you can plan, and 'careless' is not a name.
Is the gap in the content, the vocabulary, the reading load or the way the question is asked?
Why ask it
A student can know the science and lose the point on the word 'justify'. Read the item aloud to two or three students, or reword it, and see whether the answer changes. Ask this first for word problems in math and for students who are still learning the language of the test.
What did the students who got it right do that the others did not?
Why ask it
Set a strong paper beside a weak one. The difference is frequently a visible step that can be taught outright: marking up the text, drawing a diagram, checking units. If nothing on the page separates them, ask a successful student to talk you through it.
Did students run out of time, skip questions or stop writing partway?
Why ask it
Blanks bunched at the end suggest pacing or stamina. Blanks scattered through the test suggest a type of question students avoid. Neither is the same as a wrong answer, and reteaching content will not fix either one.
Is this the same mistake we saw in the last unit, or with last year's students?
Why ask it
An error that comes back every year with new children points at the curriculum or at the handoff from the grade below, and not at this class. Pull last year's notes if someone kept them. If nobody did, start a one-line record per unit today.
If we asked students why they missed this, what would they say?
Why ask it
Then go and ask three of them, having each one talk through a missed item while you listen without correcting. Their account can overturn the team's theory in five minutes. Some tests may not be shown to students afterward, so ask what the rules are for the one you gave.
Next steps
Which of these causes are within our control?
Why ask it
Draw two columns on the board: home life, last year's teacher and the bell schedule in one, and lessons, grouping, feedback and the use of class time in the other. The meeting should stay in the second. A team that lingers in the first leaves with nothing to do on Monday.
What will we reteach to the whole class, what to a small group, and what to one or two students?
Why ask it
Let the counts decide: something most of the class missed goes back to everyone, a cluster of eight gets a small group, and two students get a conference. Write the three lists during the meeting. Reteaching all thirty a skill that six of them missed spends the other twenty-four students' time.
How will the reteaching be different from the first time?
Why ask it
'The same lesson, slower' is the answer to push back on. Ask for another way in: a concrete model, a worked example with the common error left in for students to find, a different text. If a colleague's class did better, the difference is probably sitting in their plan.
On which days will this happen, and what comes out of the schedule to make room?
Why ask it
A plan without a date tends not to happen. Name the days and the minutes, then name what gets shortened or dropped, because something has to. If nothing can give, take that to whoever sets the pacing instead of agreeing to a plan that cannot fit.
Who is doing what before we meet again?
Why ask it
Each action gets one name and one date, and someone reads the list back before people stand up. The warning sign is a list where every line begins with 'we'.
What will we tell students about their results, and what will they do with them?
Why ask it
A score on its own tells a student whether to feel good or bad. Knowing which skills they have and which one comes next gives them something to aim at. Choose the form now, whether a short conference, a tracker or a corrections task, and keep each student's results out of classmates' sight.
Do families need to hear about this, and what would be useful for them to know?
Why ask it
A parent can use a message that says what the student does well, the one thing being worked on, and when you will check again. Ask how your school expects results to be shared and who makes the contact. Otherwise a family hears from three teachers, or from none.
If we could make only one change, which would it be?
Why ask it
Data meetings produce long lists that fade by Wednesday. Forcing a single choice shows what the team really believes the cause was. If people cannot agree on it, that disagreement is worth more of the meeting than the remaining agenda.
Checking back
How will we know the reteaching worked, and what will we use to check?
Why ask it
Decide the check now: three or four fresh items on the same skill, a short written task, a two-minute conference. Do not reuse the original questions, or you measure memory of the test. A small check next week tells you more than waiting for the next big assessment.
What result would tell us it did not work, and what would we try then?
Why ask it
Set the bar before the check, for instance nine of the twelve students doing it unaided, so a small bump cannot be declared a success afterward. Agree on the second move as well. It might be a different adult, a different setting, or your school's process for more intensive help.
When will we look at these same students again?
Why ask it
Put the date in the calendar before the meeting ends, close enough that the unit is still fresh. Open that meeting with the names from this one. Without that, every meeting starts from zero and nobody finds out whether anything helped.
What should change in how we teach this unit next year?
Why ask it
Write it where next year's team will trip over it, in the unit plan itself and not in the minutes. Two lines are enough: the misconception to expect and the lesson that needs an extra day.
Did this assessment give us what we needed, and what would we change about it?
Why ask it
If the team spent the meeting guessing why students missed things, the test had too few items per skill or too little shown work. For your own assessments, rewrite it now while the gaps are obvious. For one set by a district, a state or a publisher, ask what item-level reports exist and how to get them.
Since our last data meeting, what did we actually do differently in class?
Why ask it
Ask it at the start of the next meeting, before any new numbers appear. If the honest answer is 'nothing', spend less of the hour on analysis and more on planning, or meet less often. Looking at results only earns its time when a lesson changes because of it.
How to run a student data meeting with these questions
Practical guidance for the conversation itself
Before the meeting
Pick one assessment and one thing you want to find out
A meeting that tries to cover the benchmark, the unit test, grades and attendance for every class ends with a tour of the numbers and no plan. Choose one recent assessment, or one group of students, and write the question you want answered at the top of the agenda. Grades and attendance come in as supporting evidence for those same students, which is how the questions under Students use them.
Bring the work as well as the scores
You need four things on the table: a copy of the assessment, results by item and not only by student, a handful of actual student papers at different levels, and the attendance record for the weeks the unit ran. The questions under Errors cannot be answered from a score sheet. If the test was given on a platform, find out beforehand which item-level reports it can produce.
Send the results ahead
Reading a spreadsheet together for the first time uses up the minutes meant for deciding what to do. Share the results a day early and ask each person to arrive with one thing they noticed and one prediction that turned out wrong. That covers the first questions under The data before anyone sits down.
Agree on how student names are handled
This list asks for names, because a plan needs them. Keep named results inside the team that teaches those students, collect printouts at the end, and leave names out of anything displayed or sent more widely. What may be shared, with whom and in what form depends on where you work, so ask your school or district what its rules are.
In the meeting
Take the groups in order
The tags follow the way the conversation needs to go: The data, Students, Teaching, Errors, Next steps, Checking back. Skipping ahead is the usual failure. A team that starts at Next steps is planning from a hunch, and a team that never leaves The data has admired the problem. Nobody needs every question on the page in one sitting, so take two or three from each group.
Describe first, explain second
Keep the opening minutes to statements anyone could check against the page. Park every 'because' on a side list. When you reach Teaching and Errors, go back to that list and test each guess against the student work. Some will hold. Be strictest with the guesses about effort, which are the easiest to make and the hardest to find in a student's paper.
Go from the percentage to the name to the paper
A percentage tells you there is a problem, a name tells you whose it is, and the student's paper tells you what it is. Each step down gets closer to something a teacher can act on. If the meeting has stayed at percentages for more than a quarter of its time, pick up a paper.
Protect the last third for the plan
Set a timer if you have to. In a 45-minute meeting, that leaves about fifteen minutes for Next steps and Checking back, which is barely enough to settle who reteaches what, on which day, and how you will check. Analysis expands to fill whatever time it is given, and the plan is the part that reaches students.
Make it safe to show a weak result
Teachers will only bring real numbers and real papers to a room where a low score is treated as information. Whoever leads can go first with a class result of their own that disappointed them. Comparing classrooms is for finding out what worked, and the moment it is used to rank people, the useful data stops arriving.
After the meeting
Leave with one page
The record of a data meeting fits on a page: the students named, the skill each group is working on, who is teaching it, on which days, and the check that will follow with its date. Longer minutes rarely get read. Keep the page where the team can open it at the start of the next meeting.
Start the next meeting with this one
Before any new results appear, go back through the names and ask what happened. Did the reteaching take place, what did the check show, and who still needs something else? This is the habit that turns separate meetings into a cycle, and it is why the last group of questions is on the list.
Close the loop with students
Students are the people with the most use for the results and often the last to see them in a usable form. A two-minute conversation that names what a student has learned and the one skill that comes next does more than a returned score. Their own explanation of an error is also evidence the team could not get any other way.
Mistakes to avoid
Explaining results with things about the students
Motivation, home life and 'this group' may all be real, and none of them is something the team can change by next week. An explanation that leaves the teaching exactly as it was is not finished. Push for one more step: given that, what will we do differently in the lesson?
Reading a trend into a few students
Small groups swing widely from one test to the next, and a single class is a small group. Ask for counts, look at more than one assessment, and hold big conclusions until a pattern repeats. Two children having a bad day is not a reason to rewrite a unit.
Watching only the students near the line
When the measure is how many pass, attention drifts to the students who are closest to passing. Those far below and those well above then go a year without anyone planning for them. The questions on growth and on students who already have the skill are there to pull the meeting back.
Stopping at the analysis
A meeting can be thoughtful, well run and entirely about why the scores are what they are, and change nothing. If people leave without a name, a day and a check, the students who were discussed will get the same lessons as before. Cut the analysis short before you cut the plan.