Friday, June 7, 2013


Science Rules

The rules of science have never been voted on by science in general to my knowledge.  But any systematic investigation of nature and knowledge might be recognized as science if several principles are observed.

Definition of terms is a good starting point.  When terms are agreed upon study of the properties of the subject of investigation can benefit by the accurate sharing of knowledge.  It allows knowledge of the subject to accumulate when it is recorded and dispersed among participants.

Review of the literature is essential to avoid needless duplication of effort beyond the duplication needed to assure the information is correct.  The geometric growth of information has brought the need to get the latest and the best.   The latest is not always the best.  Reading science journals and attending scientific meetings often provokes one into thinking of a further aspect needing research.

Objectivity is needed by scientists.  For a long time, scientific writing was phrased in stilted expressions to avoid the use of the pronoun, “I”.   The hope was that the investigator would be less likely to have a personal agenda that would skew the results.  Truthfulness and honesty are expected of all scientists, regardless of their style of writing.

Methods and materials should be chosen as appropriate for the field of study.  The review of the literature is relied upon to help guide choices.  The materials and methods of laboratory and field researchers may be quite different, especially in biology.

Observations and data accumulated by scientific investigators are the basic materials of science.

Hypotheses are tentative explanations or principles for the results of a study.  They can be promoted to theories or laws if they accurately predict results of additional observations.  A hypothesis should be viewed in competition with all reasonable alternative hypotheses.

Experimentation in its simplest form is manipulation of one treatment variable in a set of observations in the experimental group while a control group has no change or a standard treatment.  Ideally, assignment of members to the two groups should be done randomly.  Statistical analysis of the performance of the two groups can sometimes detect even small differences in results between the two groups.

Results are presented with text, tabular, and/or graphic explanations as appropriate.

Analysis or discussion of data and other aspects may be essential parts of research reports.  Flaws in logic and overlooked implications of results should be avoided if caustic complaint letters to the editor are not desired.

Publication and sharing of results of a scientific study is expected by peers in the discipline.  Typically, results are presented to peers in meetings or some other way of sharing before publication occurs.

Peer review is used to filter information published in leading science publications.  It is done insure accuracy and validity of work submitted for publication.  When peer review is working effectively science benefits.  Some failures of peer review can curtail scientific progress when manuscripts submitted from competing schools of thought are not well evaluated.

Predictability is an implied property of science.  The prime prediction, if is is science it should be repeatable. Other outcomes may be suggested by the results of research.  

For example, my analysis of data of others led me to the probable prediction that life in the abyss of the ocean is extended far beyond other's expectations.  I used the assumption as follows: an excellent study was done of deep-sea brittle stars by Dr. Frank Rokup. His sampling was done quarterly over a year and analyzed as was reasonable for annual changes.  I combined his published data in to one sample, thinking the slow pace of life in the abyss made them equivalent to an instantaneous sample.  The result gave a size distribution similar to some long-lived invertebrates in coastal waters where predation was low.  This seemed to me to confirm my view of extreme age for some abyssal animals.  The graphic results were presented in the last chapter of Invertebrate Zoology, 3rd ed. (Engemann and Hegner, 1981).

Some problems afflict the scientific enterprise

To some scientists, data reigns supreme.  Analysis of data sometimes misses consideration of a reasonable alternative hypothesis.  Methods may not be appropriate.  Precision of technological aspects can exceed application of reason.

Reason got a bad name in science when “armchair” scientists reached their conclusions by just thinking about something when data to answer the question was easily obtained.  Publication outlets and granting agencies often have policies where editorial policy and/or peer review eliminates consideration of research if it does not involve collecting original data.

Meta-analysis and review articles typically do not contain new data.  They are generally done be recognized experts in the field.  Meta-analysis may generate new data from existing data, typically from numerous similar original studies.  That is about as close as mainstream science gets to armchair science. 

The, publish and get a grant or perish, competition in academia has to some extent spurred efforts by scientists.  But the result is often multiplication of publications with small gains and great overlap.  Graduate students can be so important as co-authors to a faculty member’s success that research specialization may interfere with breadth of training.  Production of graduate students may bring rewards to faculty so much that the screening and training of candidates may suffer.  Grade inflation occurred as pleas of those needing educational deferments during wars following World War II seemed to be silently endorsed by faculty.  The overall result of such responses may raise questions about the product of academe.

I intend to have numerous blogs about studies I found flawed.  But that is the way science works.  We get answers that are accepted until a better answer comes along.  Sharing them is a way of seeing how things might be improved.  Even a Nobel Prize winner will make my list of forthcoming blogs.  Is someone out there ready to blog about errors of my views?

Joseph G. Engemann     May 24, 2013  

 Note below added June 7, 2013

I just posted a blog on peer review.  While looking through some old reprints on the peer review process I ran across a review that raised the question, is biology science?  Some of the key qualities of science is that it is repeatable and potentially refutable.  Being repeatable means description of materials and methods is essential if verification is desired.  Laboratory scientists think the controlled experiment is the ultimate in science.  Where does that leave the field biologists whose work is largely descriptive?  One of the reprints raised the question that it might be more philosophy than science.

I don’t subscribe to that view.  The methods of the field biologists do not often have easily described controls.  Their controls may be the norms of other field observations.  Details of the environment can be reduced to controlled experiments, but it then raises the question of what the responses would be in the variables of the natural environment.  When field and lab research overlaps, it is difficult for each side to fully appreciate the contributions of the other side.


No comments:

Post a Comment