Data quality assurance

Lab Note #20
May 09, 2014
Hi again!  Maria is still actively measuring isotopes in the UW IsoLab.  Why is it taking soooo long?  First of all, she has about 300 samples.  And you've seen how our measurement requires multiple steps, which can be quite time consuming.  At this stage of the game, we are most concerned about the quality of our data.  So how do we know the quality of our data?  In other words, how do we know that the data that the mass spectrometer spits out represents reality?  This is done in two ways.

1) Along with our samples, we measure "standards".  "Standards" are essentially nitrate of known isotopic composition.  We use 4 different standards of known isotopic composition, bottles of which are pictured below.  These standards are internationally recognized, and are used by the world-wide community that performs these types of measurements for calibration and data quality purposes.  We purchase these standards from the International Atomic Energy Agency (IAEA), some of which are made by the IAEA and others are made by the United States Geological Survey (USGS) - hence the labels on the brown bottles shown below.  We run about 40 samples a day, but really only 24 of these are actual samples.  The other 16 are a combination of the four standards shown below.  We are continually checking to see if we are getting accurate values.

2) We measure each sample 3 times.  This way we get a sense of our precision.  It also makes our lives less stressful since we know we have at least 3 tries to get it right.  What can go wrong?  Well, lots of things since there are lots of steps. The most common problem is the bacteria.  Although we try really hard to make them happy and healthy, they are living things, and sometimes they aren't so healthy.  An unhealthy batch of bacteria can create isotope havoc, which we can easily diagnose by checking our four isotope standards.  We use the triplicates to calculate the precision, which show up as error bars on our plots of the data.

Accuracy and precision, that is the goal!

Maria is finishing up round two, and will present her data at our group meeting next week.  I will let you know if the data look as expected.

Happy Mother's Day!

Please wait...