In Inspect data for harvest time we used a simplistic approach to find the harvest time. Since you have replicates it is possible to use a more rigorous statistical approach.
Student's t-test
The Student's t-test is a great way to check if the cell concentrations differ significantly for the data points close to max cell titer. A seemingly higher arithmetic mean may simply be due to inherent variability. The goal of the t-test is to identify the first data point from where cell concentration is no longer increasing significantly.
You can use the Excel template below to perform a t-test on your data. Although this step is optional and will likely lead to the same conclusions as the step-by-step guide above, it offers a more rigorous statistical basis for your analysis and can increase confidence in your decisions. The Student's t-test is the rigorous approach to the above (simplistic) visual inspection of overlapping error bars.
To do this, you will need to export the data by using the EXPORT CSV button in the measurement group panel.
Example of data export from Access to CSV file. The CSV file is opened in a spreadsheet tool such as Excel.
Admittedly, it is tedious to export data from Access and analyze it in Excel. That is why we are working on curve-fitting features directly in Access. All the tricky math will happen automatically and you get a user-friendly and rigorous approach to choose the best harvest time.
Stay tuned for updates!
We are working on a harvest algorithm
The most rigorous approach to determine harvest time is to fit the data to a mathematical function. In the present workflow we use a more simplistic, rule-of-thumb approach.
Stay tuned. We are working on a when-to-harvest algorithm directly in Access. It will do all the tricky curve-fitting automatically.
