Counting only truly independent replicates

Science · Scientific Inquiry · ages 22-23

Name ______________________ Date ____________

A specimen is measured repeatedly under the same conditions. You want one summary per assigned unit. What is a suitable first step?
- Count each repeat as a new independently assigned unit
- Average suitable repeats within each assigned unit
- Delete unit identifiers before pooling all the repeats
Which practice is pseudoreplication?
- Treating dependent readings as independent units
- Repeating a study with new independently assigned units
- Recording repeated readings with their unit identifiers
Two separately prepared cell cultures share one donor. Separate culture-level treatment assignments do not make them independent donor-level replicates.
Circle one: True False
Treatments are assigned independently to patients, and each patient has repeated readings. Which analysis respects this design?
- Pool readings and count each as a separately assigned patient
- Keep only the patient with the greatest number of readings
- Account for patient units and dependence within each patient
Readings have positive correlation within assigned units. What can happen if a standard-error calculation treats every reading as independent?
- It can underestimate the uncertainty of the mean
- It can make n too small and overstate uncertainty
- It can increase the actual number of independent units
A plot shows 15 readings from five independently assigned animals. Before interpreting its error bars, what should you check?
- Whether there are enough readings for an independent-data formula
- What the bars represent and how animal grouping was handled
- Whether averaging readings creates more independent animals
Four classes are independently assigned teaching methods. Each class has 20 measured students. Which count describes the treatment assignments?
- 80 assignments, one for every student who was measured
- 20 assignments, matching the number of students per class
- 4 assignments, one for each independently assigned class
Accounting for within-unit dependence leaves the effect estimate unchanged but increases its standard error. The result is no longer significant. What follows?
- The estimated effect became smaller after the analysis changed
- Greater estimated uncertainty changed the significance decision
- The change in uncertainty indicates that more units were collected

Answer key

For grown-ups. Fold this page away before handing over the rest.

Counting only truly independent replicates W1-mt_2gRWB6RTaI-s1

Average suitable repeats within each assigned unit · A within-unit mean supplies the requested summary without pretending that repeats are new assignments.
Treating dependent readings as independent units · The problem is ignoring dependence in the analysis, not collecting or recording repeated measurements.
True · There is one donor, despite the two culture-level assignments. The cultures are not two independently sampled donors.
Account for patient units and dependence within each patient · Patient summaries or a suitable hierarchical model can respect the patient-level assignment and repeated readings.
It can underestimate the uncertainty of the mean · The independent-observation calculation omits positive within-unit dependence and can make the estimated standard error too small.
What the bars represent and how animal grouping was handled · The plotted point count does not reveal the error-bar calculation. A valid analysis may show all readings while accounting for animals.
4 assignments, one for each independently assigned class · Methods are assigned to whole classes. The 80 student measurements occur within four assigned classes.
Greater estimated uncertainty changed the significance decision · The estimate stayed the same. Its greater uncertainty changed the threshold decision, without proving a zero effect.