Tags

, , ,


Are you a person who loves making decision supported by good numbers?  Then be careful before jumping to your decision, you have to listen carefully – what those numbers are trying to say?. Today is 25th anniversary of Columbia Shuttle disaster. For me it is reminder for a 25 year old story about pre-launch data analysis done by NASA engineers – check details at Wikipedia “Space Shuttle Challenger disaster” – who justified the launch without paying attention to the fact that test data for some critical components was outside the extreme cold temperature range, which exist at the time of launch.

Challenger explosion (image borrowed from wikipedia)

We do also look at data everyday, practically most of the time we do look summary generated by set of samples – either stock quotes, weather reports, component quality at shop floor, consumer confidence index, or test result of software testing in lab. We all know that it is not always possible or sometime not even practical for individual to analyze raw data himself. We are bound to rely on other expert analysis reports, which summarize description of data.  But as a consumer of those reports, you are responsible to have clear understanding of data sample source, reliability of sample collection, assumptions made in analysis, and principals applied in preparing the report.  Just reading numbers and making your interpretation can misguide you like a half painted picture.

If decision is critical you may also want to check out for outliers and anomalies which were ignored in compiling summary. If you are looking at comparison for two reports, just be sure that they follow same mechanism in collecting data and follow same principals and assumptions from analyst. Otherwise you would end-up comparing  apple with oranges.

So next time whenever you would be making a critical decision based on data, be sure that you had listen to those numbers before making your judgement and you are sure that you clearly understand what those numbers were saying to you.

Advertisements