Don’t rely on fancy statistics - Deepstash

Bite‑sized knowledge

to upgrade

your career

Ideas from books, articles & podcasts.

Don’t rely on fancy statistics

Don’t rely on fancy statistics

They can obscure the structure of evidence.

Remain skeptical of all statistical and mathematical modeling procedures that pre-process evidence while leaving little trace of its original structure.  

Start out close to the data by scrutinizing frequency distributions, means, standard deviations, how variables are actually measured, degree of measurement error, zero-order correlations, and sample composition.

1

STASHED IN:

112

MORE IDEAS FROM THE SAME ARTICLE

Ignore the author’s conclusions.

When you are doing cross-disciplinary research what really matters is the structure of the full body of evidence rather than any authors’ particular interpretation of their data in one paper, which is often biased. 

It will force you to reexamine the full body of evidence with new eyes.

Seeking relevant knowledge in unfamiliar disciplines increases the odds of serendipitous insights and allows novel empirical patterns to emerge.

Be open to have your basic presumptions shattered.

Strategies for solving scientific mysteries:  

  • pin down why perspectives on them differ and what each has to offer
  • try to locate the essential core of available evidence
  • not put m...

Go on cross-disciplinary research expeditions.

By reading and translating the literature in fields outside your own, the full body of evidence surrounding a problem become apparent. 

Discover and save more ideas by creating a

FREE

Deepstash account.

Develop a

reading habit

, save

time

and create an amazing

knowledge library

.

GET THE APP:

MORE LIKE THIS

Treating Failure Like a Scientist

When a scientist runs an experiment, there are all sorts of results that could happen: Some are positive and some are negative, but all of them are data points. Each result is a piece of data that can ultimately lead to an answer.

3

STASHED IN:

69

Statistics Inaccuracies

Most studies that involve statistical research remain largely inaccurate, and a large number of hypotheses use data samples which are inadequate.

1

STASHED IN:

50

Most traffic jams are unnecessary. Local transport engineers manage to achieve local improvements, but after a while the flows rearrange, and the same traffic jams appear elsewhere.

Mathematicians who specialize in traffic flow would like to solve urban traffic jams forever.

STASHED IN:

49