By Madeline McCurry-Schmidt / ASAS Communications
Think you’re Sherlock when it comes to inspecting data? Debra Aaron, a professor in the department of animal sciences at the University of Kentucky, says a better understanding of statistics can help students answer the big questions behind their research.
Aaron thinks too many students wait until after they’ve finished their experiments to think about how they are going to analyze data and interpret statistics. She says that taking courses in statistics before designing experiments can help students stay organized and consider the data as they go.
“Oftentimes, we’re doing exploratory types of data collection and analysis,” Aaron said.
By understanding what data is really important to an experiment, students can make sure their experiments actually address their hypotheses.
She said she has seen cases where students devote time to experiments that aren’t set up to gather the right kinds of data. Maybe the population size is too small or the control group isn’t set up right. If students understood how their data will be used after an experiment, they might design better experiments, Aaron says.
“Somehow they don’t see stats as as important as a laboratory methods course,” Aaron said.
With many animal scientists relying on computer programs to process data and produce statistics, Aaron says it is important for students to also try doing calculations by hand.
“Otherwise you have no idea what that computer program is giving you,” Aaron said.
Aaron said many students also work with university statisticians to analyze results after they finish experiments. She recommends meeting with statisticians before even starting an experiment.
“You don’t go see the statistician when it’s too late,” Aaron said. “Oftentimes, the statistician gets stuck telling students really bad things.”
Understanding how statistics work can help student researchers spot the right clues in their data.
“With statistics, you change one thing and the whole story changes,” Aaron said. “So having those types of courses will make you a better detective.”
Pingback: What NOT to do in cancer research | ASAS Graduate BULLetin