Science Hub · Primary Science · PSLE Science Learning Guide · Reality Lab
Real-world science often arrives as a label, graph, headline, image, rating, test result or percentage. The Reality Lab teaches you to ask what the evidence actually supports before you accept the claim.
This is the canonical index for the eduKate Sengkang PSLE Science Reality Lab. Each volume applies Primary 5/6 scientific inquiry and evidence reasoning to one real-world communication object. Reality Lab does not replace the core PSLE Science guides on variables, fair testing, graphs, measurement, conclusions or scientific explanation; it gives you real situations in which to use those skills.
How to use the Reality Lab
- Choose a claim, metric or representation that looks familiar.
- Before opening the article, say what you think the number, image, label or headline proves.
- Check what was actually measured, compared, estimated or inferred.
- Notice the boundary between the evidence and the stronger claim someone might be tempted to make.
- Try a different volume and see whether the same reasoning habit transfers.
Complete Reality Lab index
Index verified through Vol.354. This hub lists the public canonical Reality Lab volumes in numerical title order. Unreleased volume numbers are intentionally absent.
Volumes 001–100
- Vol.001 — “Scientifically proven”: what was actually tested?
- Vol.002 — “Twice as effective”: compared with what?
- Vol.003 — A testimonial is not yet a fair test
- Vol.004 — What evidence does a causal headline need?
- Vol.005 — “95% agree”: who was actually asked?
- Vol.006 — A demonstration can show what happened without testing why
- Vol.007 — What happens when the time window changes?
- Vol.008 — Three websites are not necessarily three independent pieces of evidence
- Vol.009 — “Every trial worked”: were failures left out?
- Vol.010 — Does a lab result still hold outside the laboratory?
- Vol.011 — “No germs detected” does not prove there were none
- Vol.012 — “Biodegradable”: under what conditions and how long?
- Vol.013 — “Energy-saving”: were both devices doing the same job?
- Vol.014 — “Up to 70% better”: best result or typical result?
- Vol.015 — “Waterproof”: how deep, how long and under what conditions?
- Vol.016 — A simulation prediction is not the same as an observation
- Vol.017 — More foam is not automatically better cleaning
- Vol.018 — “Certified”: what exactly was checked?
- Vol.019 — “Recyclable”: recyclable where, and what happens after the bin?
- Vol.020 — Company funding does not by itself decide whether evidence is true
- Vol.021 — A bigger icon does not necessarily mean a bigger scientific effect
- Vol.022 — Are before-and-after pictures really comparable?
- Vol.023 — 100 samples can still come from one batch
- Vol.024 — “50% recycled material”: 50% of what?
- Vol.025 — “Eco-friendly”: which environmental result was measured?
- Vol.026 — A corrected study: what changed and what still stands?
- Vol.027 — Peer reviewed does not mean proven correct
- Vol.028 — What did one overall score hide?
- Vol.029 — What happened to the missing results?
- Vol.030 — Did a second axis create the apparent pattern?
- Vol.031 — What did averaging hide behind a smooth trend?
- Vol.032 — Did colour boundaries create the map pattern?
- Vol.033 — Does an upward line describe the individual results?
- Vol.034 — A red satellite image does not mean the forest is literally red
- Vol.035 — Was the prediction written before the results were known?
- Vol.036 — “Live data”: when did the dataset actually stop updating?
- Vol.037 — One composite picture may combine observations from different times
- Vol.038 — A map value does not mean that exact point was directly measured
- Vol.039 — A gap-free map may contain filled-in values
- Vol.040 — A pixel’s quality flag matters as much as its number
- Vol.041 — A 30°C forecast is not one certain future
- Vol.042 — “Blind tested”: what did the tester not know?
- Vol.043 — What contamination did a clean blank actually check?
- Vol.044 — A model fitting the data does not prove the explanation
- Vol.045 — “95% accurate”: accurate on which data?
- Vol.046 — If both groups improved, did the product cause the change?
- Vol.047 — Three agreeing studies may share the same dataset
- Vol.048 — Did the measuring method change where the trend jumps?
- Vol.049 — If a test found nothing, did a positive control show it could detect something?
- Vol.050 — Traceable to a national standard does not mean product endorsement
- Vol.051 — Is the laboratory accredited for this exact test?
- Vol.052 — “Randomized test”: what was actually random?
- Vol.053 — One successful test: how many others were tried?
- Vol.054 — Last month’s calibration does not guarantee today’s reading
- Vol.055 — How was a “representative image” chosen?
- Vol.056 — Observed every two days is not continuous observation
- Vol.057 — The ± uncertainty matters to the claim
- Vol.058 — “2°C warmer than normal”: normal compared with which years?
- Vol.059 — “Warmest on record” gives rank, not the size of the difference
- Vol.060 — 10,000 sightings do not mean people looked equally everywhere
- Vol.061 — Near-real-time data may not be the final science-quality result
- Vol.062 — 30 m resolution: one point or a whole patch?
- Vol.063 — A vegetation-health index is not a direct measurement of plant health
- Vol.064 — “Data adjusted”: what changed and why?
- Vol.065 — Open data does not mean the result has already been reproduced
- Vol.066 — Did image enhancement reveal a feature or create an artifact?
- Vol.067 — Within tolerance does not mean exactly on target
- Vol.068 — Is a result robust to other reasonable analyses?
- Vol.069 — Time-lapse playback can change the apparent rate
- Vol.070 — Three agreeing sensors can share the same bias
- Vol.071 — 10,000 accelerated cycles do not automatically prove years of normal use
- Vol.072 — “Removes 99%”: 99% of what starting amount?
- Vol.073 — Normalised lines starting at 100 can hide different real starting values
- Vol.074 — Were both products dropped the same way?
- Vol.075 — Did an experiment video cut out anything between steps?
- Vol.076 — “Stops growth” may overstate a result that only showed slower growth
- Vol.077 — Bigger-looking cells: were the images at the same scale?
- Vol.078 — A truncated axis can exaggerate bar differences
- Vol.079 — A pooled result does not describe every sampled location
- Vol.080 — An average passing a limit does not mean every item passed
- Vol.081 — Compare failure counts only after checking denominators
- Vol.082 — A rising total does not necessarily mean the process is speeding up
- Vol.083 — Red can mean different values on different maps
- Vol.084 — A bigger slice can come from a smaller whole
- Vol.085 — “Lasts twice as long”: did both tests use the same endpoint?
- Vol.086 — Where do measurements end and projections begin?
- Vol.087 — “Absorbs twice as much”: were equal amounts tested?
- Vol.088 — “30% less material”: per pack or per same number of uses?
- Vol.089 — “Highest ever”: ever since when?
- Vol.090 — A flatline at 100 may be a sensor limit, not a stopped process
- Vol.091 — Did the first trial change the starting conditions for the second?
- Vol.092 — Extra decimal places do not guarantee extra precision
- Vol.093 — “No added substance Q” does not prove the final product contains none
- Vol.094 — “Detected” does not tell you how much is there
- Vol.095 — Can a value of 140 be trusted when the calibration curve stops at 100?
- Vol.096 — Could something else trigger the same sensor signal?
- Vol.097 — 1, 10, 100, 1000 can be equal steps on a logarithmic axis
- Vol.098 — “Tested in triplicate”: three independent samples or three readings of one sample?
- Vol.099 — Plant-based plastic does not automatically mean biodegradable
- Vol.100 — Can a corrected negative value represent a negative physical amount?
Volumes 101–200
- Vol.101 — Same score, same measurement and test?
- Vol.102 — Same number, same units?
- Vol.103 — Filtered before testing: does the result describe the original sample?
- Vol.104 — Moisture content: wet basis or dry basis?
- Vol.105 — Doubled concentration does not prove twice as much substance was added
- Vol.106 — Holding time: what happened before testing?
- Vol.107 — QA/QC passed does not mean every result is correct
- Vol.108 — Duplicate samples disagreed: should they simply be averaged?
- Vol.109 — Spike recovery 95% is not 95% removed
- Vol.110 — “Method validated” needs sample, range and purpose
- Vol.111 — Can a reporting limit of 10 support a claim below 5?
- Vol.112 — An estimated result is not the same kind of number as an ordinary measured value
- Vol.113 — Dilution changes how the original-sample reporting limit is interpreted
- Vol.114 — Same timestamp, same time zone?
- Vol.115 — A map projection may not preserve area
- Vol.116 — “No statistically significant difference” does not prove sameness
- Vol.117 — Tiny error bars do not automatically mean accuracy
- Vol.118 — Matching labels and chain of custody
- Vol.119 — Retesting only the lowest results can create an apparent improvement
- Vol.120 — Still working at test end does not make lifetime equal to test length
- Vol.121 — Process in control is not the same as product meeting specification
- Vol.122 — What does a 30% chance of rain mean?
- Vol.123 — A 100-year flood does not mean 99 safe years follow
- Vol.124 — p = 0.04 is not a 96% chance the claim is true
- Vol.125 — A 95% confidence interval is not where 95% of individual results must fall
- Vol.126 — A reading inside a limit can still require uncertainty awareness
- Vol.127 — >100 is not exactly 100
- Vol.128 — Retraction does not prove the opposite claim
- Vol.129 — Doubled AQI does not necessarily mean doubled pollution
- Vol.130 — Below detection limit is not automatically zero
- Vol.131 — Homogenising a sample can hide local variation
- Vol.132 — Missing mass is not proof it was destroyed
- Vol.133 — A wider forecast cone does not mean a bigger storm
- Vol.134 — A preprint has not automatically passed peer review
- Vol.135 — Certified reference value does not mean every tiny scoop is exact
- Vol.136 — Passing a proficiency test does not guarantee every future result
- Vol.137 — Check the averaging period behind a displayed value
- Vol.138 — Two methods can disagree without one automatically being wrong
- Vol.139 — Blank map cell does not automatically mean zero
- Vol.140 — A positive count depends on the cutoff threshold
- Vol.141 — A verified outlier is not the general rule
- Vol.142 — Brighter microscopy image does not automatically mean more target
- Vol.143 — 1,000 camera-trap photos are not 1,000 animals
- Vol.144 — Artist’s impression is not a telescope photograph
- Vol.145 — Red weather radar does not guarantee that rain intensity at the ground
- Vol.146 — “Validated data” does not mean every value is true and final
- Vol.147 — A pooled forest-plot diamond does not mean every study agrees
- Vol.148 — Ensemble forecast lines are not different storms
- Vol.149 — Rainfall 20 mm is not automatically 20 mm of standing water
- Vol.150 — “Tested to a standard” can mean method or pass requirement
- Vol.151 — Accuracy specification is not the uncertainty of one reading
- Vol.152 — River gauge height is not depth everywhere
- Vol.153 — eDNA detection does not prove a live animal was at that exact spot
- Vol.154 — Accepted lot does not mean every item was tested
- Vol.155 — A box-plot outlier is not automatically wrong
- Vol.156 — R² = 0.90 is not 90% of predictions correct
- Vol.157 — 100% inspected does not guarantee every defect was found
- Vol.158 — Fire hotspot pixel does not mean the whole square is burning
- Vol.159 — Reanalysis maps are not direct measurements at every grid cell
- Vol.160 — Satellite NO₂ column is not ground-level concentration
- Vol.161 — Same earthquake magnitude does not mean same shaking everywhere
- Vol.162 — Significant wave height is not the height of every wave
- Vol.163 — Sea-surface temperature is not the whole ocean depth
- Vol.164 — Hurricane category does not encode size, rain and surge
- Vol.165 — “Feels like” heat index is not a thermometer reading
- Vol.166 — Sea-ice concentration is not ice thickness
- Vol.167 — An epicentre dot still has location uncertainty
- Vol.168 — Night-time brightness is not a population count
- Vol.169 — Wind gust is not sustained wind speed
- Vol.170 — Relative humidity 80% does not mean 80% of air is water
- Vol.171 — Turbidity 20 NTU is not 20 particles
- Vol.172 — Global mean sea-level rise is not the same change at every coast
- Vol.173 — PM2.5 mass concentration is not particle size or count
- Vol.174 — Precipitable water 40 mm is not a 40 mm rainfall forecast
- Vol.175 — HEPA 99.97% does not mean the room becomes 99.97% particle-free
- Vol.176 — Streamflow 100 m³/s is not river speed 100 m/s
- Vol.177 — 800 lumens is not 800 lux everywhere
- Vol.178 — 1 kg CO₂e is not necessarily 1 kg of CO₂ emitted
- Vol.179 — Water hardness “as CaCO₃” is an equivalent reporting basis
- Vol.180 — A 400 W solar-panel rating is not 400 W all day
- Vol.181 — 60 dB is not twice the sound intensity of 30 dB
- Vol.182 — LED colour temperature 3000 K is not physical bulb temperature
- Vol.183 — Salinity 35 PSU is not 35% salt
- Vol.184 — Total ozone 300 DU is not a 300 mm-thick layer
- Vol.185 — Sea-level pressure may be reduced from the station reading
- Vol.186 — Wind direction 270° means from the west, not toward the west
- Vol.187 — Specific conductance 500 µS/cm is not 500 mg/L salt
- Vol.188 — Snow water equivalent 100 mm is not 100 mm snow depth
- Vol.189 — Visibility 10 km is not a hard invisible wall
- Vol.190 — Thermal-camera 50°C is not necessarily a direct surface-temperature measurement
- Vol.191 — CADR 300 m³/h does not mean a 300 m³ room becomes completely clean in one hour
- Vol.192 — Octane 95 is not 95% octane
- Vol.193 — Cooling capacity 12,000 Btu/h is not electricity input
- Vol.194 — Battery capacity 5000 mAh is not a fixed current or runtime
- Vol.195 — 350 kWh/year is a comparison estimate, not guaranteed home use
- Vol.196 — 100 CFU/mL is not an exact cell count
- Vol.197 — Water activity 0.85 is not 85% water
- Vol.198 — 80% dissolved-oxygen saturation does not mean 80% of the water is oxygen
- Vol.199 — BOD₅ 5 mg/L is not 5 mg/L of bacteria
- Vol.200 — Insulation R-value 4 is not four times the heat blocked
Volumes 201–300
- Vol.201 — Heat-pump COP 3.5 is not energy creation
- Vol.202 — MERV 13 is not 13% particle capture
- Vol.203 — 12 °Brix is not automatically 12% sugar
- Vol.204 — Six air changes per hour does not mean every molecule is replaced six times
- Vol.205 — UV Index 8 is not 8% UV
- Vol.206 — GPS accuracy 5 m is not a guaranteed 5 m error bound
- Vol.207 — pH 5 vs 6 is not a small linear step
- Vol.208 — Albedo 0.30 is not 30% white
- Vol.209 — Dew point 20°C is not a forecast low
- Vol.210 — Solar irradiance 1000 W/m² is not panel output
- Vol.211 — Secchi depth 2 m is not lake depth
- Vol.212 — Volcano WATCH is not a certain eruption
- Vol.213 — Capacity factor 30% is not 30% efficiency
- Vol.214 — CRI 90 is not 90% colour accuracy
- Vol.215 — Mohs 8 is not twice as hard as Mohs 4
- Vol.216 — “99.9% pure” needs a defined basis
- Vol.217 — Specific gravity 3.0 compares equal volumes
- Vol.218 — EF3 wind speed may be inferred rather than directly measured
- Vol.219 — Earth-sized exoplanet is not another Earth
- Vol.220 — Organic molecules detected is not life found
- Vol.221 — High landslide susceptibility is not a forecast
- Vol.222 — Habitable zone does not guarantee habitability
- Vol.223 — Suitable habitat does not mean species present everywhere
- Vol.224 — Tensile strength 500 MPa is not a whole-object load
- Vol.225 — High tide 1.8 m is height relative to a datum, not water depth
- Vol.226 — IP68 is not a universal waterproof guarantee
- Vol.227 — L70 50,000 h is not sudden LED failure at hour 50,000
- Vol.228 — Wildfire 80% contained is not 80% extinguished
- Vol.229 — Pasteurized is not sterile or automatically shelf-stable
- Vol.230 — 90% germination does not mean exactly nine of every ten seeds
- Vol.231 — MTBF 100,000 h is not device lifetime
- Vol.232 — Population index 120 is not 120 animals
- Vol.233 — Carbon neutral is not zero emissions
- Vol.234 — Potentially hazardous asteroid is not an impact prediction
- Vol.235 — Exoplanet candidate is not a confirmed planet
- Vol.236 — “Endangered” does not tell you the exact population count
- Vol.237 — D3 Extreme Drought is not one measurement or a forecast
- Vol.238 — RPD 5% is not 95% accuracy
- Vol.239 — AQL 1.0 is not a guarantee of at most 1% defects
- Vol.240 — Tree-ring reconstruction is not direct temperature measurement
- Vol.241 — Halved concentration does not necessarily mean half the river load
- Vol.242 — Emission factor is an estimate, not necessarily a direct source measurement
- Vol.243 — Site occupancy is not the observed detection fraction
- Vol.244 — Satellite chlorophyll-a is not a direct phytoplankton count
- Vol.245 — Satellite soil moisture may describe a shallow surface layer, not the root zone
- Vol.246 — One grab sample is not the whole river all day
- Vol.247 — A climate normal is not a daily forecast
- Vol.248 — Displayed digits do not guarantee last-digit accuracy
- Vol.249 — Correlation r = 0.80 is not causation
- Vol.250 — An air sensor is not automatically equivalent to a regulatory monitor
- Vol.251 — A two-hour fire rating is not a real-fire stopwatch
- Vol.252 — Cpk 1.33 does not guarantee every product
- Vol.253 — Cycle life 500 does not mean dead at cycle 501
- Vol.254 — A hindcast match does not guarantee a future forecast
- Vol.255 — Passive-sampler average is not constant concentration
- Vol.256 — Pilot-scale 90% removal is not a full-scale guarantee
- Vol.257 — Detected in 70% of samples is not 70% of time or area
- Vol.258 — Aerosol optical depth is not ground-level PM2.5 concentration
- Vol.259 — Land-surface temperature is not air temperature
- Vol.260 — Species richness is not an individual count
- Vol.261 — Fire Weather Index is not wildfire probability
- Vol.262 — Battery State of Health is not current charge
- Vol.263 — First-pass yield is not final good output
- Vol.264 — Canopy cover is not tree count or total forest health
- Vol.265 — System availability is not power output or capacity factor
- Vol.266 — Screening-level exceedance is not proof harm has occurred
- Vol.267 — Population estimate is not a literal animal count
- Vol.268 — Percent increase needs the starting baseline
- Vol.269 — Energy-label ticks compare relative efficiency, not total use
- Vol.270 — Satellite rainfall estimate is not a rain gauge at every point
- Vol.271 — RMSE 2°C does not mean every forecast is within 2°C
- Vol.272 — Zero waste to landfill is not zero waste produced
- Vol.273 — Higher diversity index is not simply more species
- Vol.275 — Industrially compostable is not backyard compostable
- Vol.276 — Round-trip efficiency 90% is not charge retention
- Vol.277 — ISO 9001 does not mean every product was tested
- Vol.278 — T90 response time is not 90% accuracy
- Vol.279 — SEER2 20 is not 20% efficiency
- Vol.280 — STC 50 does not mean every sound is 50 dB quieter
- Vol.281 — Rated EV range is not guaranteed trip distance
- Vol.282 — eDNA read share is not animal-abundance share
- Vol.283 — Microplastic particle count is not mass concentration
- Vol.284 — Coral bleaching alert does not mean all coral is dead
- Vol.285 — Tree-cover loss is not automatically permanent deforestation
- Vol.286 — Zero-VOC label does not mean zero emissions
- Vol.287 — Doppler radial velocity is not full wind speed
- Vol.288 — Pocket TDS meter may use conductivity as a proxy
- Vol.289 — Forest carbon stock is not annual sequestration rate
- Vol.290 — River stage is not river depth or discharge
- Vol.291 — Data sonification is not a recorded sound in space
- Vol.292 — 100 W equivalent is not 100 W power input
- Vol.293 — UPF 50 is not 50% UV blocked
- Vol.294 — Black-hole image reconstruction is not an ordinary photograph
- Vol.295 — Sensor sensitivity is not percent accuracy
- Vol.296 — 9H pencil hardness is not Mohs 9
- Vol.297 — Treadwear 400 does not guarantee four-times real-road life
- Vol.298 — Higher CPUE does not prove there are more fish
- Vol.299 — R² = 0.99 is not 99% accuracy
- Vol.300 — Geiger-counter CPM is not dose rate
Volumes 301–400
- Vol.301 — Salt-spray hours vs real-world service life
- Vol.302 — OD600 is not an exact cell count
- Vol.303 — SPAD 40 is not 40% chlorophyll
- Vol.304 — RSD 2% is not 98% accuracy
- Vol.305 — 1000× magnification is not 1000× more detail
- Vol.306 — Base peak 100 is not 100% sample composition
- Vol.307 — Noise dose 100% is not 100 dBA
- Vol.308 — WBGT 30°C is not simply air temperature
- Vol.309 — Heating degree days are not a count of cold days
- Vol.310 — CO₂ 800 ppm is not 800 mg/m³
- Vol.311 — Transit depth 1% is not planet width
- Vol.312 — Leaf Area Index 3 is not three leaves or 300% cover
- Vol.313 — Cloud optical thickness 20 is not 20 km thickness
- Vol.314 — Sea-surface height anomaly +10 cm is not a global 10 cm rise
- Vol.315 — What a 10-10-10 fertilizer label actually means
- Vol.316 — Water-storage anomaly is not groundwater-table drop
- Vol.317 — Kp 7 is not seven times Kp 1
- Vol.318 — 80% biobased content is not 80% of whole-product mass
- Vol.319 — NDVI 0.8 is not 80% vegetation cover
- Vol.320 — A 10,000 ha wildfire perimeter does not mean all 10,000 ha burned
- Vol.321 — A plastic resin code is not a recycling guarantee
- Vol.322 — “Kills 99.9% of germs” is not every germ in every use
- Vol.323 — 500 recorded calls are not 500 animals
- Vol.324 — A 99% DNA-barcode match is not automatic species certainty
- Vol.325 — “Control = 100%” is a reference, not absolute units
- Vol.326 — 70% biodegradation is not 70% visible disappearance everywhere
- Vol.327 — ENERGY STAR score 75 is not 75% efficiency
- Vol.328 — Fire Radiative Power 50 MW is not whole-fire output
- Vol.329 — 10th-percentile streamflow is not 10% of normal flow
- Vol.330 — NSF/ANSI 53 certification does not mean every contaminant is removed
- Vol.331 — SPF 50 does not mean exactly 50 times longer in the sun
- Vol.332 — Preregistration is not proof a result is correct
- Vol.333 — A correction notice does not make a whole study invalid
- Vol.334 — UL Listed is not a never-fail guarantee
- Vol.335 — False-colour satellite imagery is not literal colour
- Vol.336 — HEPA 99.97% at 0.3 µm does not mean 0.3 µm is the smallest captured size
- Vol.337 — Magnitude 7 is not seven times magnitude 1
- Vol.338 — ISO/IEC 17025 accreditation has a defined scope
- Vol.339 — 48 MP does not guarantee four times the real detail
- Vol.340 — A 1C battery charge rate is not an exact one-hour charge
- Vol.341 — Water resistant 10 bar is not permission for any 100 m activity
- Vol.342 — NRR 30 dB is not an exact personal 30 dB reduction
- Vol.343 — A water year is not necessarily a calendar year
- Vol.344 — Exoplanet equilibrium temperature is not measured surface temperature
- Vol.345 — CAPE 3000 J/kg is not a severe-storm guarantee
- Vol.346 — Drag coefficient 0.30 is not 30% of the force
- Vol.347 — Ocean heat-content anomaly +100 ZJ is not 100°C hotter
- Vol.348 — A global satellite mosaic is not one same-moment photograph
- Vol.349 — 22% solar-panel efficiency does not mean all the rest is waste heat
- Vol.350 — Beaufort Force 8 is not 8 m/s
- Vol.351 — 100 MPGe does not mean one gallon of gasoline was used
- Vol.352 — LAeq 70 dB is not a constant 70 dB sound level
- Vol.353 — D50 = 20 µm does not mean half the particles are exactly 20 µm
- Vol.354 — Baseflow Index 60% is not 60% groundwater at every moment
Return to the PSLE Science learning system
Reality Lab is the application lane. For the underlying skills—question reading, evidence selection, variables, fair testing, graphs, measurement, models, conclusions, explanation, checking and revision—return to the PSLE Science Learning Guide.
