How to Organize a School Science Fair with STEM Toy Kits | Planning Guide for Teachers

Define the Project Type

Before students choose a kit, make sure everyone is clear about what kind of project they are doing. A science investigation, an engineering project, and a demonstration may all use the same STEM kit, but the work students need to produce is different.

For an engineering project, students need a clear target before they start building. That target might limit the amount of material, size, weight, time, or cost. NGSS engineering standards use criteria, constraints, testing, comparison, and redesign as core engineering practices.^[1]

If students may enter an ISEF-affiliated fair, do not force an engineering project into a standard hypothesis format. Society for Science separates research questions and hypotheses from engineering goals in its project-planning rules.^[2]

Before approving a project, check that the student can clearly answer:

• what they are testing or designing;
• what they will measure;
• how they will know whether the design worked;
• what data they will keep;
• what they will change if the first version fails.

Set the Fair Format

The format should match the equipment you actually have. If the class has fewer kits than students, forcing everyone into an individual project usually creates long waits and very little hands-on time.

Individual projects make sense when students have enough equipment, projects are not competing for the same test area, and the teacher has enough time to review separate proposals.

Team projects are usually a better fit when robots, circuits, sensors, or other kits are limited. Students can split the work across building, programming, testing, and recording, but each person should still understand the full project.

For 30 students and only 6 kits, five-student groups are possible. Another option is to run two testing rotations with smaller groups so each student gets more time with the equipment.

Station projects work well when the school owns several different types of kit. Students can try different areas before choosing one for the fair.

When several groups share electronic kits, keep a basic record of:

• program version;
• gear or wheel setup;
• sensor position;
• battery status;
• missing parts;
• known damage.

That prevents one group from starting with a device that has already been changed by someone else.

Class challenges work best when everyone follows the same overall brief but makes their own design choices.

Example: every group receives 30 building pieces and must create the most stable tower. The material limit stays the same; the design does not.

Screen the Kits

A kit does not need to look advanced. It needs to give students room to make a change and then measure what happened.

Before using a kit, ask:

• What can I change?
• What can I measure?

A robot kit might let students change wheel size, gear ratio, code, or sensor position, then measure distance, time, accuracy, or successful runs.

A construction kit might let students change structure shape, base width, beam arrangement, or gear layout, then compare load, height, stability, or speed.

Science kits need a different kind of check. A science activity kit may need one lesson, several days of observation, consumable materials, storage, or extra cleanup. Those practical details matter before a teacher assigns the project.

Students get more value from a kit when they can build something, test it, make a change, and test again. Piano Potato’s overview of hands-on STEM learning kits shows several ways this type of activity can be used.

Age and safety: Read the manufacturer’s age label and warnings first. U.S. toy requirements cover areas including battery-operated toys, magnets, mechanical hazards, and other toy-specific risks.^[3]

Classroom difficulty: Two products with similar age labels may still be very different in practice. One may need basic assembly; another may require much more reading, planning, or problem-solving. This is particularly relevant with building STEM toys, where part count and design freedom can change how much support students need.

Replacement parts: Check whether you can replace:

• motors;
• wheels;
• axles;
• gears;
• wires;
• sensors;
• connectors;
• battery holders.

A kit is much harder to manage in a classroom if one lost or broken part takes the whole set out of use.

Software: Check this before buying, not after students start the project.

• Does it need an app?
• Does it work on school devices?
• Does it require internet access?
• Does it require student accounts?
• Can students save different program versions?
• Can the project still be explained if the software fails on fair day?

Check Class-Time Fit

A kit can look suitable on paper and still be a poor fit for a normal lesson if setup and reset take too long.

For a 45-minute period, one workable division is:

If a kit needs 25 minutes just to assemble or reset, only 20 minutes remain for everything else in a 45-minute class.

Reset time can make a big difference during testing. In a 20-minute testing block:

• a 3-minute reset can support several test cycles;
• a 10-minute reset may allow only one or two useful cycles.

Not every project should be forced into a single class period. A crystal-growing kit, for example, works better with fixed observation times over several days than with repeated tests in one class.

Approve Projects Before Testing

Students should not start testing as soon as they choose a topic. A short proposal catches weak questions and unrealistic plans before time and materials are wasted.

Science proposal:

• question;
• prediction;
• one main factor being changed;
• result being measured;
• conditions being kept similar;
• materials;
• test method;
• planned number of trials;
• safety concerns.

Engineering proposal:

• problem;
• design goal;
• success rule;
• design limits;
• first design;
• test method;
• measurement;
• safety concerns.

Revise the proposal if:

• the school does not have the measuring tool;
• the test cannot be repeated;
• the student cannot complete the main work;
• too many things change at once;
• the project cannot fit the available schedule;
• the test creates unnecessary risk.

ISEF-affiliated projects require a research plan, and some projects require review before experimentation begins.^[4]

Lock the Test Conditions

The easiest way to get confusing data is to change more than one important thing without noticing it. STEM kits have several hidden variables that are easy to miss.

Robots:

• battery charge;
• wheel alignment;
• starting position;
• floor surface;
• gear connection;
• program version.

Sensors:

• sensor angle;
• sensor distance;
• room lighting;
• calibration.

Solar models:

• light source;
• distance from light;
• panel position;
• shadows.

Structures:

• material quantity;
• span;
• load position;
• connector tightness;
• previous damage.

Testing order matters too. If all large-wheel tests happen at the end, a weaker battery may make that setup look worse than it really is.

Weak order:

• small wheels × 5;
• medium wheels × 5;
• large wheels × 5.

Better check:

• small;
• medium;
• large;
• small;
• medium;
• large.

Alternating the conditions reduces the chance that only the last setup is affected by changing equipment performance.

Set the Trial Count

For a simple classroom comparison, 3–5 trials can be a practical starting point. If the results are all over the place, do not stop just because the planned number of trials is complete.

Four trials can still produce very different levels of consistency:

Set B needs more checking because the results vary much more.

If one result looks unusual, do not remove it just because it is inconvenient. Record a clear reason if something went wrong, such as the robot hitting a chair. If there is no clear reason, keep the value and discuss the variation.

Choose the Right Data

If a robot succeeds in 8 of 10 maze runs:

8 ÷ 10 × 100 = 80%

Program A is faster. Program B is more reliable. The better choice depends on what the student decided to optimize before testing.

Keep units consistent. Do not mix centimeters and inches in the same data set.

Do not add decimal places the measuring tool cannot support. A basic classroom ruler does not justify a result such as 12.4837 cm.

Students do not need an average for every project. Maximum load, success rate, number of failures, or completion time may be more useful.

If no clear difference appears, write:

No clear difference was found under these test conditions.

Build the Schedule

With two 45-minute science periods each week for 8 weeks, students have:

8 × 2 = 16 class periods

Week 1: Rules, project types, kits, safety.

Week 2: Topic and proposal.

Week 3: Final method and material check.

Weeks 4–5: Build, test, retest.

Week 6: Data table, graph, limitation, next test.

Week 7: Display, repair, missed tests, software check.

Week 8: Setup, safety check, judging, public session, cleanup.

Keep the catch-up period free until you actually need it. A missing student, broken sensor, weak data set, or software problem can quickly use up that time.

Build the Budget

The sticker price of the kit is only part of the cost. Batteries, spare parts, measuring tools, and display materials also need a place in the budget.

Illustrative 30-student classroom budget:

$500 ÷ 30 = about $16.67 per student

If the reusable kits remain available the following year:

• $50 replacement parts;
• $50 consumables;
• $70 display materials.

$170 ÷ 30 = about $5.67 per student

These are planning examples, not market averages.

Keep spare parts for the small items that can stop an entire project:

• charging cables;
• battery holders;
• connectors;
• wheels;
• axles;
• wires.

Set Parent and Student Boundaries

Parents can support the project without taking ownership of it.

Parents can:

• help transport materials;
• help find simple supplies;
• supervise tools when required;
• help with typing when appropriate;
• ask practice questions.

Parents should not:

• design the project;
• invent or replace data;
• write the conclusion;
• complete the main build;
• answer the judge for the student.

If a student is not allowed to use a tool, the student should still make the design decision. The student can mark the required cut or hole; an authorized adult can operate the tool.

Use normal school accommodations for students who need:

• larger parts;
• simpler written instructions;
• adapted controls;
• extra setup time;
• a different presentation method.

Run the Safety Check

NSTA recommends school science safety programs that address hazards, training, supervision, facilities, and appropriate procedures.^[5]

Check every project for:

• batteries;
• electricity;
• magnets;
• small parts;
• sharp parts;
• moving mechanisms;
• tools;
• heavy loads;
• heat;
• chemicals or liquids;
• projectiles.

For each risk, answer three questions:

1. What can go wrong?
2. How will we control it?
3. Can it operate during the public fair?

Batteries: Use the battery type and charging method specified for the kit. Remove damaged, swollen, leaking, or unusually hot battery packs from use.

CPSC warns that swallowed button or coin batteries can cause severe internal injury and that children should not have access to loose batteries or unsecured battery compartments.^[6]

Magnets: Keep loose high-powered magnets controlled. CPSC warns that swallowed magnets can attract each other through internal tissue and cause severe injury.^[7]

Moving projects: Do not add sharp blades, unsafe projectiles, exposed high-speed parts, or modifications that remove protective covers.

Human testing: If students ask people to test an app, prototype, invention, or device and the project may enter an ISEF-affiliated fair, check the human-participant rules before testing begins.^[8]

Map the Room

Work out the project count before arranging the room. If 60 projects are displayed at 2 projects per table position:

60 ÷ 2 = 30 project-table positions

That does not include judging areas, robot courses, check-in, storage, or visitor space.

Static display area:

• posters;
• models;
• non-moving demonstrations.

Robot area:

• marked floor route;
• no path toward visitors;
• no path toward stairs or fragile displays.

Structure area:

• keep falling loads away from feet;
• do destructive tests before public viewing when needed;
• ask where the load will fall if the model fails.

Judge area:

• enough room to stand beside the project;
• away from loud robot or mechanical tests where possible.

Any project that depends on a working device should also have a backup:

• photos;
• video;
• raw data;
• diagrams;
• program screenshots.

ISEF display rules restrict some materials at booths and allow photographs or video to represent items that cannot be physically displayed.^[9]

Score Projects

Science Method: variables, controlled conditions, measurement, repeatability.

Engineering Method: design goal, limits, test method, redesign.

NGSS middle-school engineering standards emphasize comparing solutions against criteria and constraints and using test results to improve designs.^[10]

Example score:

With presentation worth only 5 points, a polished display cannot make up for weak testing.

Calibrate the Judges

Before formal judging, give every judge the same sample project and ask them to score it independently.

If one judge gives 92/100 and another gives 58/100, the rubric is being interpreted differently and needs to be discussed before students are scored.

Keep the judge questions consistent:

• What did you change?
• Why did you measure that?
• What did you keep the same?
• Which result surprised you?
• What evidence supports your conclusion?
• What would you test next?

For 60 projects at 8 minutes each:

60 × 8 = 480 project-interview minutes

With 8 judges working at the same time:

480 ÷ 8 = 60 minutes

Leave extra time for walking between projects, writing scores, delays, and judge discussion.

Build the Display

Science project:

1. Question
2. Prediction
3. Materials
4. Method
5. Data
6. Result
7. Conclusion

Engineering project:

1. Problem
2. Design Goal
3. Limits
4. First Design
5. Test Results
6. Changes
7. Final Design

Keep the raw data available so judges can see where the final numbers came from.

Result: The 6 cm wheels traveled an average of 103 cm.

Explanation: One possible reason is that this wheel size gave the motor a better balance between wheel rotation and travel distance.

The result comes from the data. The explanation is still an interpretation and should not be written as if the project proved it directly.

Prepare the Student Talk

Science:

• My question was…
• I changed…
• I measured…
• The data showed…
• One limitation was…
• Next time I would…

Engineering:

• The problem was…
• My design had to…
• I tested it by…
• The first design failed because…
• I changed…
• The final design…

Students should keep conclusions inside the limits of their own data.

Avoid:

Large wheels are always better.

Use:

In my tests, the 6 cm wheels produced the greatest average distance.

Limitations should also be specific. “Human error” does not tell the judge much.

Better examples include:

• battery charge changed;
• a wheel loosened;
• room lighting changed;
• only one surface was tested.

Use Ready-to-Run Project Models

Robot wheels: Test 4 cm, 6 cm, and 8 cm wheels. Four trials per size gives 3 × 4 = 12 runs.

Ramp angle: Test 10°, 20°, 30°, and 40°. Four trials per angle gives 4 × 4 = 16 runs.

Solar panel angle: Test 0°, 30°, 60°, and 90°. Three readings per angle gives 4 × 3 = 12 readings.

Robot maze: Compare two programs with 10 runs each. Total: 20 runs. Record both time and success rate.

Bridge design: Test three designs with the same material amount, bridge span, and loading method.

Gear ratio: Test three gear arrangements with the same motor, wheel, battery, and test time.

Earthquake structure: Set one material limit, one minimum size, one fixed test time, and one success rule. Require at least one redesign.

Rock and mineral comparison: Use known specimens to compare hardness, streak, color, or texture instead of only displaying the samples. A fixed rock and mineral collection can keep the specimen set consistent across groups.

Catch Weak Projects Early

If five bridges all hold the maximum test load of 2 kg, the test may be too easy to compare the designs.

If all five fail at the first 200 g load, the test may be too difficult.

These numbers are examples only; teachers should use safe levels suited to the materials and age group.

Handle Failed Tests

Device failure: A motor, battery, cable, or sensor stops working. Record the equipment problem.

Test failure: A bridge collapses or robot leaves the course. Keep the result if it happened during a valid planned test.

Prediction failure: The result does not match the student’s prediction. Keep the data. The prediction does not need to be correct.

If the method changes during the project, record:

• the original method;
• the problem;
• the new method;
• the date of the change.

Teacher Checklist

• Fair date confirmed
• Room confirmed
• Kits counted
• Shared-kit schedule ready
• Software tested
• Chargers ready
• Spare parts ready
• Projects approved
• Safety checks complete
• Judges assigned
• Moving-project areas marked
• Backup photos or video available
• Accessibility needs checked
• Kit return plan ready

Student Check

• What exactly did I test or design?
• What evidence supports my conclusion?
• What was hardest to keep the same?
• What limitation affected the result?
• What would I test or change next?

FAQ

What if two students choose the same question?

Use different designs, materials, test conditions, or measurements. Each student should collect their own data.

What if a kit breaks during testing?

Record the replacement. Do not automatically combine data from different motors, sensors, or devices if they may behave differently.

Can the question change after testing starts?

Yes. Record the original question, the reason for the change, and the date. Use earlier data only if it still answers the new question.

What if the results show no clear difference?

Write that no clear difference was found under the tested conditions. Do not invent a stronger conclusion.

What if a robot or app fails on fair day?

Use saved data, photos, diagrams, program records, or test video.

Finally

For a 30-student fair, 10 shared kits can support 10 teams of 3. An 8-week schedule with two 45-minute classes per week gives 16 periods; keep about 6 for building and testing and 1 as a repair or retest buffer. Start simple comparisons with 3–5 trials, then add tests when results vary widely. Use a 100-point rubric that puts most points on method, data, analysis, and understanding. Before fair day, every student should be able to explain one result, one limitation, and one next test without reading from the display.