Too little
Selecting the desired result
“Leave that point out; the story is clearer without it.”
The desired conclusion supplies the reason for choosing which evidence to show.
Report the observation, including what changed.
Dispositional refusal to mislead through what one says or represents.
Honesty represents what you know truthfully. In science, it preserves the actual observation, relevant conditions and distinction between a result and an expectation. Other people should be able to understand the work from the record you provide.

The report is nearly ready. One observation sits far from the others; remove it, and the effect looks much more impressive. Your cursor hovers over the point as though tidiness were an analytical reason. You return to the stated criteria, keep the record intact and show how the conclusion was reached. The dot has no idea how inconvenient it is.
An awkward observation tempts an investigator to change an analysis because the preferred result would look stronger.
Too little
“Leave that point out; the story is clearer without it.”
The desired conclusion supplies the reason for choosing which evidence to show.
A fitting response
“We will apply the stated criteria and explain any exclusion or later analysis.”
The reader can distinguish the planned account, justified corrections and exploratory work.
Pursued too far
“An honest report must treat even a known recording error as a valid measurement.”
Refusing justified correction can misrepresent the evidence just as surely as convenient deletion.
Honesty preserves the basis of the decision. A correction needs a reason that survives its effect on the conclusion.
A faithful report gives another investigator something they can examine rather than merely admire. They can judge the observation, the decision rule and the interpretation separately. A less striking result may still sharpen the question or prevent wasted effort. Clear records also make later correction possible without requiring anyone to reconstruct a vanished analysis.
Two weights differ because someone took a bite between readings. Record the changed sample beside the values.
An interesting pattern suggests a new comparison. Label that exploration clearly instead of presenting it as the original prediction.
Intellectual Honesty governs what you allow yourself to believe. Honesty governs how faithfully you represent what you know.
Candor makes the consequential truth plain; Honesty requires the underlying representation to be truthful.
Calling an inconvenient point “noise” after seeing its effect disguises selection as routine care. Examine it under defensible criteria and preserve the reason for the decision.
A report can display many technical details while omitting the one analytical choice that changes the finding. Include the consequential choice where readers can understand it.
Preserve the original observations and the reasons for changes.
Distinguish the planned analysis from work suggested by its results.
Explain consequential exclusions and corrections in the report.
Review one ordinary summary you own. Check whether its selection or correction of information has a reason you can state independently of the outcome you prefer.
Truthful reporting can respect privacy and confidentiality. Use an appropriate record or channel rather than exposing unrelated personal information.
Seven selections to return to through the week. Follow a source link to read the surrounding passage.
Recorded observation consists of two distinct parts: 1st, an exact notice of the thing observed, and of all the particulars which may be supposed to have any natural connection with it; and, 2dly, a true and faithful record of them.
John F. W. Herschel · Preliminary Discourse on the Study of Natural Philosophy · § 112
Herschel separates noticing from faithfully recording. An inconvenient observation must be accounted for in the report; seeing it privately does not make its unexplained absence informative to a reader.
One of its numerous processes is to make multitudes of observations, and out of these to select those only which agree, or very nearly agree.
Charles Babbage · Reflections on the Decline of Science in England · Chapter V, Section 3
Babbage calls this selective reporting “cooking.” The problem is choosing observations for the agreement they produce; an independently justified exclusion is a different decision and needs an explicit record.
What happens next, in an uncertain light, does not lend itself to accurate observation.
Jean-Henri Fabre · The Wonders of Instinct · Chapter 10 · trans. Alexander Teixeira de Mattos
Fabre marks a limit in what he could observe. A record becomes more trustworthy when it says where the available view was inadequate.
TRIMMING consists in clipping off little bits here and there from those observations which differ most in excess from the mean, and in sticking them on to those which are too small; a species of "equitable adjustment," as a radical would term it, which cannot be admitted in science.
Charles Babbage · Reflections on the Decline of Science in England · Chapter V, Section 3
The smoothing Babbage describes changes observations to make a preferred pattern. A faithful graph can remain untidy; its appearance does not supply a reason to alter the record.
The corrections of the errors of the instrument by means of little screws are given up, and the errors which are known to exist are corrected in the computations.
Maria Mitchell · Maria Mitchell: Life, Letters, and Journals · Chapter VI
Mitchell describes a historical instrument whose known faults are accounted for in calculation. Acknowledged correction differs from presenting an imperfect instrument as flawless.
… fact and reason are the ultimate arbiters, and patience and honesty are the great helpers out of difficulty.
Thomas Henry Huxley · Science and Education · Scientific Education: Notes Of An After-Dinner Speech
Huxley places honesty inside the work of resolving a problem. A truthful account of what happened makes correction possible when a polished story would hide the difficulty.
It might be expected that the same compression should result from each, if the different methods of observation could be made without error.
Mary Somerville · On the Connexion of the Physical Sciences · Section VI.
Somerville introduces disagreement among estimates of the earth’s shape rather than erasing it. Reporting the mismatch leaves room to examine the methods and their errors.
Choose this approved collection as an optional workbook program, or select individual virtues in My Virtues.