Reproducibility

Scientific experiments must be reproducible. Any other researcher should be able to replicate the experiment and get similar results.

Scientific results must also be dependable. When working in a lab, one encounters many sensitive reagents, and any contamination or changes may affect your results drastically. If you cannot vouch for every ingredient and its concentration in a reaction, you cannot conclude anything based on those results.

This is also the reason why most tests and measurements are done in triplets (three identical reaction tubes for each condition being tested). This way, if similar results are obtained from all three samples, there is a greater chance that results are dependable.