LECTURE 2 Induction: facts acquired through observation. Observing similar situations over time and seeing a common property. Problems with induction include that it is not logical to prove anything by repeated observation and it is unclear how many observations are necessary. Only takes one example to disprove theory. Deduction: predictions and explanations (logical arguments). Hyptheticodeductive approach: theories (hypotheses) are disproved because proof is logically impossible. Based on falsifcationism. Hypothesis: the prediction that relates to the theory being tested. Needs to be specific. Falsificationism: we attempt to falsity a null hypothesis (hypothesis we seek to falsify). If we reject the null we can accept the alternate hypothesis. Example of scientific method: Observation: female spiders are larger than male spiders Specific model: full grown female spiders always larger than male spiders of the same species. Hypothesis: we predict that the average mass of female spiders will be greater than males for each species of spider encountered. H1: Females > Males Null hypothesis: H0: Females Males Experiment: Gather data on as many spider species as possible Results: You find 50 species for which the prediction is true. One for which it is not (European water spider). Thus, on average, females are larger. Interpretation: We reject (falsify) our null hypothesis because we have shown that on average females are bigger than males Uncertainty (variability): measurements from the natural word are variable. Uncertainty caused by: natural variation among individuals (place, time, genotype, individual, population, communities) or items being measured and error in measurement. Our challenge is to detect patterns against a background of naturally occurring variation. Variable: an attribute of biological systems that we are interested in, measure, and can vary from individual to individual. Sample: collect information from a subset of the population. Population: collection of all possible measurements Sample size (n): The number of observations per sample Random sample: is one where all individuals in the population are equally likely to be in your sample. If samples are random, or nearrandom, then, average (mean) density should represent the population density. Random is not only valid way of sampling: Systematic sampling: selected according to a random starting point and a fixed, periodic interval. Systematic with random elements: random with defined zones Haphazard: does not follow any systematic way of selecting participants. Stratified random: involves the division of a population into smaller groups known as strata. Biased sampling: biased samples are more likely to be represented. LECTURE 3 (listen later) There are 2 main types of data and studies: Observational (mensurative): no active treatments applied to the objects of interest in the study. Not usually referred to as an experiment. Eg. Weather patterns, health trends. Limitation: Association does not equal causation. Manipulative: you allocate subjects to a treatment, manipulating the study. The treatment is our manipulation and the control group is our reference point. Dependant variable: subject of interest in the study (X). Independent variable: variable thought to be a predictor of Y. Experimental design: how to properly test a prediction. Involves avoiding confounding variables and use of replication. Many characteristics of things we think are the same vary naturally in space and time. Replication: refers to independent sampling units. Each individual should generate one datum (one observation).