Our scientific grade metering equipment is rated at +/-4% of full scale and the reptile lamp meters he referred to are +/-10% of full scale.
That breaks down to a possible error of 8% total on our instruments and a possible error of 20% on the SOLARMETER MODEL 6.2 UV METER. That is a 250% difference in accuracy!
For Peter: The Taiwan GI meter is hardly "scientific". Most such things made in Asia (not including Japan) are infamous for having labels and stickers on them saying anything they want with no documentation. Example: UL or CE approvals or ISO when none of such has been certified.
The master Solarmeters are actually calibrated ±4% ref NIST. We bump that up to ±10% on UVB label simply to account for any possible uncertainty in transferring the master reading to saleable units. In reality the saleable units read exactly same as master within 1 or 2 significant digits as shipped.
The "250% difference in accuracy" above is fuzzy relative (not absolute) math. In absolute terms... IF the actual UVB was say 400 µW/cm² then minus 10% worst case = 360. If minus 4% then 384. The 360 = 94% of 384. Therefore absolute minus difference reverts back to only 6% - not 250%! And besides - the 10% vs 4% would hardly account for the gross error of super high 11000 µW/cm² (11.0 mW/cm²) UVB as presented on the black helicopter site.