Biology · Biology in Health, Food & the Environment

Evidence, equity, access and trade-offs in biological applications

Two fictional communities each contain one hundred eligible people. Sixty in the first and forty in the second have access to an imagined application. Both have the same favourable-outcome fraction among those with access, yet different shares of the eligible population have that documented outcome. Equal conditional fractions do not guarantee equal coverage.

Research Design, Data & Scientific Practice introduces independent units, assignment, sampling and comparisons. Heredity, Variation & Evolution distinguishes inherited variants from traits and individual responses from population change. Energy, Enzymes & Metabolism introduces cellular reactions and material transformations. Their English chapters are available in this working manuscript; publication remains pending. The examples here use small graphs, fractions, sums and stated time windows, with each convention explained before use.

Back to the Biology map

By the end, you can

Compare access and outcome denominators, preserve the distribution of benefits and burdens, and distinguish descriptive evidence from a value judgement about an application.

An outcome among users does not describe everyone

Community One records sixty people with access and forty without; Community Two records forty with access and sixty without. Among the people with access, forty-eight and thirty-two have the defined favourable outcome. The application and outcome are fictional, with no medical intervention or clinical meaning assigned.

The favourable-outcome fractions among people with access are forty-eight divided by sixty and thirty-two divided by forty. Both equal four fifths. Using all eligible people instead gives forty-eight divided by one hundred and thirty-two divided by one hundred. These are shares with a documented favourable outcome after access, not estimates of the communities' overall health.

People without access have no application-outcome observation in this account. That does not mean they have no favourable outcomes in life, nor that the application caused the recorded outcomes among users. Group composition and other processes remain unspecified. The table describes access and documented outcomes while leaving efficacy and wider health effects open.

Equal user fractions coexist with different documented coverage

A conditional fraction and an eligible-population share have different denominators. This descriptive account supplies no application effect or overall population-health estimate.

Read the diagram in words

Community One: one hundred eligible, sixty with access, forty without; forty-eight documented favourable outcomes and twelve other access records. Community Two: one hundred eligible, forty with access, sixty without; thirty-two documented favourable outcomes and eight other access records. The conditional fractions are forty-eight sixtieths and thirty-two fortieths, both four fifths. Eligible documented shares are 0.48 and 0.32; outcomes without access are unknown.

Equal user fractions coexist with different documented coverage
Community	Eligible people	People with access	People without access	Documented favourable outcome after access	Access without the documented favourable outcome	Fraction among people with access	Documented share among all eligible people	Outcomes without access unknown
Community One	100	60	40	48	12	0.8	0.48	?
Community Two	100	40	60	32	8	0.8	0.32	?

Documented share among all eligible people: 0.48 − 0.32 = 0.16

Original descriptive fictional access and documented outcome records, with no actual application, clinical outcome or causal comparison. Outcomes among people without access are not observed. Population shares count only documented outcomes after access; they do not describe total health.

Totals can hide where consequences fall

A second, separate model compares two imagined schemes across two communities. Scheme First assigns ten benefit units to each community and burden values two and eight. Scheme Second assigns eight benefit units to each and burden values four and four. Benefit and burden use different scales; subtracting them would require an additional justified valuation.

Scheme First has a larger total benefit, twenty rather than sixteen, and a larger total burden, ten rather than eight. Its burden is also more unevenly distributed. For Community One it offers more benefit and less burden than Second; for Community Two it offers more of both. An aggregate comparison misses these different positions.

Equal numerical burdens do not by themselves establish equity. Communities may have different needs, baseline risks or ability to bear a cost, none of which the model specifies. Equity concerns how consequences and opportunities relate to such circumstances. Scientific accounts can clarify consequences, while a judgement about acceptable distribution also involves values and the people affected.

Aggregate gains and community burdens answer different questions

A larger total benefit can coexist with a larger and more uneven burden. Equal numerical burdens alone do not settle whether the distribution is equitable.

Read the diagram in words

First gives benefit ten and ten with burdens two and eight: totals twenty and ten. Second gives benefit eight and eight with burdens four and four: totals sixteen and eight. First minus Second adds two benefit units in each community, lowers One's burden by two and raises Two's by four. Burden ranges are six and zero. No net score is assigned.

Scheme First
Community	Assigned benefit units	Assigned burden units on a different scale
Community One	10	2
Community Two	10	8

Sum across the two communities
Scheme	Assigned benefit units	Assigned burden units on a different scale	Community distribution · Assigned burden units on a different scale
Scheme First	20	10	8 − 2 = 6

Scheme Second
Community	Assigned benefit units	Assigned burden units on a different scale
Community One	8	4
Community Two	8	4

Sum across the two communities
Scheme	Assigned benefit units	Assigned burden units on a different scale	Community distribution · Assigned burden units on a different scale
Scheme Second	16	8	4 − 4 = 0

First minus Second
Community	Assigned benefit units	Assigned burden units on a different scale
Community One	2	-2
Community Two	2	4

No shared net valuation supplied: ?; Baseline needs and capacity to bear burdens are unspecified

Separate original fictional scheme comparison. Benefit and burden have different assigned scales with no shared valuation, probability, clinical meaning, resource cost or universal equity criterion.

An application claim needs both scope and comparison

Health, food and environmental applications connect mechanisms to human purposes. A proposal can have a plausible mechanism yet uncertain human outcomes. It can improve one material account yet shift a burden elsewhere. It can have similar recorded outcomes among users while leaving access uneven. These are different questions requiring different comparisons.

Social and environmental context can shape exposure, vulnerability and opportunities to use an application. Evidence about distribution therefore complements an average. To interpret a result, identify the outcome, population, denominator, time and boundary, along with whether the comparison addresses causation or only an association. Missing information should remain visible.

A scientific model can show which conclusions follow from stated assumptions and which remain open. It does not choose every goal for a community. A transparent evaluation distinguishes measured consequences from expected ones and from value judgements. Revising a judgement when evidence or priorities change is part of evaluating an application, rather than a failure of biological reasoning.

Worked example

Preserve conditional outcomes and community burdens

Among people with access, forty-eight out of sixty and thirty-two out of forty both equal eighty per cent. Among all eligible people, the recorded application-associated favourable outcomes account for forty-eight and thirty-two per cent. The sixteen-percentage-point difference in these documented shares accompanies different access fractions in this stipulated account.

This calculation does not establish that changing access would cause that difference to disappear. The same conditional fraction is supplied descriptively; it need not stay fixed under a real change in participation or context. Outcomes without access and a causal comparison of the application are not supplied.

For the separate schemes, moving from Second to First adds two benefit units in each community. It reduces Community One's burden from four to two, but increases Community Two's from four to eight. A recommendation based solely on total benefit would omit that asymmetry. An overall net score would also require explicit weights because the two scales are different.

Check your understanding

Both communities have an eighty-per-cent favourable-outcome fraction among people with access. Why are their documented population shares different?
Show reasoning
Access reaches sixty versus forty of each hundred eligible people. The documented counts are therefore forty-eight versus thirty-two. Equal conditional fractions with different access denominators produce different eligible-population shares under the supplied account.
Can twenty benefit units minus ten burden units establish that Scheme First is better than Second?
Show reasoning
No. Benefit and burden use different scales, so that subtraction has no defined valuation. It would also hide distribution: Community Two bears a larger burden under First. A judgement needs explicit criteria and context.
Why does the model not establish that equal burden numbers are equitable?
Show reasoning
It supplies no information about differing needs, baseline circumstances or capacity to bear a burden. Numerical equality is a descriptive property; equity requires a judgement in relation to those circumstances and the people affected.

What to carry forward

Keep access, outcomes and burdens tied to their populations and boundaries. Evidence clarifies an application; priorities and distribution also require explicit judgements.

References

Resources supporting this lesson’s explanations and examples.