Many methods used in #marketing_research (MR) in the 21st century have been adapted from methods developed in medical research and epidemiology in the last century.
They aim to identify both causes and cures, as we do in MR. Like #MR, they utilize experimental, quasi-experimental and observational data.
Here are some books on these subjects I've found helpful:
🔸 - Fundamentals of Clinical Trials (Friedman et al.)
🔸 - Introduction to Statistical Methods for Clinical Trials (Cook and DeMets)
🔸 - Case-Control Studies (Keogh and Cox)
🔸 - Handbook of Statistical Methods for Case-Control Studies (Borgan et al.)
🔸 - Modern Epidemiology (Rothman et al.)
🔸 - Epidemiology: Study Design and Data Analysis (Woodward)
🔸 - Spatio-Temporal Methods in Environmental Epidemiology (Shaddick and Zidek)
🔸 - Handbook of Spatial Epidemiology (Lawson et al.)**
#DataScience has been rightly criticized for mining data for correlations that prove ephemeral ("Torture the data until it confesses. Then the data recants its confession.")
However, in fairness, complete understanding of causation is not necessary to act, otherwise humans would have vanished long ago.
There are many medical conditions we do not fully understand that we can treat effectively.
Researchers need to be both rigorous and realistic.
#Statisticians often walk a sort of tightrope with rigor on one side and reality on the other. Our clients typically want quick yes-or-no answers but we need to be careful how we phrase our explanations lest we slip and fall.
Some clients are quite curious, however, and see interactions with statisticians and methodologists as an opportunity to learn. They may even have read popular books on statistical topics by Nate Silver, Philip Tetlock or other authors.
The danger there is that these books are not completely non-technical and important points can be missed or misunderstood. (Daniel Kahneman comes to mind too, and some marketing researchers seems to have misconstrued what he'd actually written in "Thinking, Fast and Slow.")
✴️ @AI_Python_EN
They aim to identify both causes and cures, as we do in MR. Like #MR, they utilize experimental, quasi-experimental and observational data.
Here are some books on these subjects I've found helpful:
🔸 - Fundamentals of Clinical Trials (Friedman et al.)
🔸 - Introduction to Statistical Methods for Clinical Trials (Cook and DeMets)
🔸 - Case-Control Studies (Keogh and Cox)
🔸 - Handbook of Statistical Methods for Case-Control Studies (Borgan et al.)
🔸 - Modern Epidemiology (Rothman et al.)
🔸 - Epidemiology: Study Design and Data Analysis (Woodward)
🔸 - Spatio-Temporal Methods in Environmental Epidemiology (Shaddick and Zidek)
🔸 - Handbook of Spatial Epidemiology (Lawson et al.)**
#DataScience has been rightly criticized for mining data for correlations that prove ephemeral ("Torture the data until it confesses. Then the data recants its confession.")
However, in fairness, complete understanding of causation is not necessary to act, otherwise humans would have vanished long ago.
There are many medical conditions we do not fully understand that we can treat effectively.
Researchers need to be both rigorous and realistic.
#Statisticians often walk a sort of tightrope with rigor on one side and reality on the other. Our clients typically want quick yes-or-no answers but we need to be careful how we phrase our explanations lest we slip and fall.
Some clients are quite curious, however, and see interactions with statisticians and methodologists as an opportunity to learn. They may even have read popular books on statistical topics by Nate Silver, Philip Tetlock or other authors.
The danger there is that these books are not completely non-technical and important points can be missed or misunderstood. (Daniel Kahneman comes to mind too, and some marketing researchers seems to have misconstrued what he'd actually written in "Thinking, Fast and Slow.")
✴️ @AI_Python_EN