Logic and Critical Thinking in the Biomedical Sciences

1. Learning what counting tells usSection 1.1. Science is mostly about counting stuff Section 1.2. Never count on an accurate count Section 1.3. Large samples cannot compensate for nonrepresentative data Section 1.4. The perils of combining data setsSection 1.5. Compositionality: Why small outnumbers large Section 1.6. Looking at data Section 1.7. Counting mutations Section 1.8. Chromosome length and the frequency of genetic diseases Section 1.9. Counting instances of species Section 1.10. Counting garbage Glossary References 2. Drawing inferences from absences of data values Section 2.1. When the important data is what you do not see Section 2.2. The power of negative thinking Section 2.3. Absence of x-rays emitted by hot cups of coffee Section 2.4. Absence of laboratory findings in SIDS (sudden infant death syndrome) Section 2.5. Absence of lethal toxicity resulting from damage to the epigenome and systems that regulate gene expressionSection 2.6. Absence of deficiency diseases among highly conserved genesSection 2.7. Absence of shared conserved noncoding elements Section 2.8. Absence of animals with built-in wheels Section 2.9. Absence of microcancers Section 2.10. Absence of frogs on small islands Section 2.11. Absence of great apes roaming outside Africa Section 2.12. Absence of penguins in northern hemisphere Section 2.13. Absence of samarium-146 isotope from earth Section 2.14. Obligation to look for absences Glossary References 3. Drawing inferences from data ranges Section 3.1. Why are data ranges important? Section 3.2. The range of dust sizes that cause human disease Section 3.3. When tumor cells have very small nuclei Section 3.4. The range of heights that animals can jump Section 3.5. Blood chemistry Section 3.6. Narrow ranges of enzyme activity Section 3.7. The number of different types of cancers Section 3.8. Limits imposed by the dynamic range of measuring instruments Glossary References 4. Drawing inferences from outliers and exceptions Section 4.1. One is the loneliest number Section 4.2. Ozone, the outlier that couldn’t be believed Section 4.3. Neoplasms having very short latency periods Section 4.4. Outliers as sentinels for common diseases Section 4.5. How exceptions elucidate pathogenesis Section 4.6. Finding the outliers Glossary References 5. What we learn when our data are abnormal Section 5.1. Creating normal distributions Section 5.2. Pareto’s principle and Zipf distribution in biological systems Section 5.3. Pareto’s bias: Favoring the common items Section 5.4. Recognizing composite diseases Section 5.5. Multimodality in population data Section 5.6. Removing some of the mystery around ovarian cancers Section 5.7. Living with Berkson’s paradox Glossary References 6. Using time to solve cause and effect dilemmas Section 6.1. Timing is everything Section 6.2. Does anybody really know what time it is? Section 6.3. Temporal paradoxes Section 6.4. Timing the progression of cancer development Section 6.5. When the temporal sequence is observed incorrectly Section 6.6. Smoke and mirrors Section 6.7. Refusing simple answers Section 6.8. Dose-dependent effects and the fallacy of causation Section 6.9. Time-window bias Section 6.10. Replacing causation with pathogenesis Glossary References 7. Heuristic methods that use random numbers Section 7.1. The value of randomness Section 7.2. Repeated sampling Section 7.3. Monte Carlo simulations for tumor growth and metastasis Section 7.4. A seemingly unlikely string of occurrences Section 7.5. Cancer is not caused by bad luck Section 7.6. Several approaches to the birthday problem Section 7.7. Modeling cancer incidence by age Section 7.8. The Monty Hall puzzle Glossary References 8. Estimations for biomedical data Section 8.1. The inestimable value of estimates Section 8.2. The limit of hemoglobin concentration in red blood cells Section 8.3. CODIS: How to do it all without having it all Section 8.4. Some useful approximation methods Section 8.5. Some useful numbers Glossary References