All statistical models have assumptions, and violation of these assumptions can affect the reliability of any conclusions we draw. Before we fit any statistical model, we need to explore the data to be sure we fit a valid model. Are relationships assumed to be a straight line really linear? Does the response variable follow the assumed distribution? Are variances consistent? We outline several graphical techniques for exploring data and introduce the analysis of model residuals as a powerful tool. If assumptions are violated, we consider two solutions, transforming variables to satisfy assumptions and using models that assume different distributions more consistent with the raw data and residuals. The exploratory stage can be extensive, but it is essential. At this pre-analysis stage, we also consider what to do about missing observations.

We don’t always have a single response variable, and disciplines like community ecology or the new “omics” bring rich datasets. Chapters 14–16 introduce the treatment of these multivariate data, with multiple variables recorded for each unit or “object.” We start with how we measure association between variables and use eigenanalysis to reduce the original variables to a smaller number of summary components or functions while retaining most of the variation. Then we look at the broad range of measures of dissimilarity or distance between objects based on the variables. Both approaches allow examination of relationships among objects and can be used in linear modeling when response and predictor variables are identified. We also highlight the important role of transformations and standardizations when interpreting multivariate analyses.

In this chapter the author addresses the following questions: What does it mean to say that a language is a creole? Do creoles constitute a separate global typological class apart from other language typological groupings? The author calls for research on creole languages that is free from linguistic feature bias, creole language list bias, and genealogical bias. He compares expansion languages from the Meso and South American indigenous language families, particularly the Quechuan family (focusing on Ecuador, the northern border), with the Arawakan, Tupian, Cariban, Jêan, Chibchan, Uto-Aztecan, Mayan, and Otomanguean families. The comparison highlights processes of ethnogenesis, morphological reduction, and sub- or adstrate influence. The findings help broaden the definition of “creole” to refer to a special lexifier– descendant relationship, making the notion of creole a relational one, to facilitate comparisons with other languages and with “linguistic areas.” Questions that remain include: Are there area-specific features? And can this approach shed light on the characterization of creoles as a particular group of languages? What role can language contact play in reconstructing language families?

Poor-quality measurements are likely to yield meaningless or unrepeatable findings. High-quality measurements are characterised by validity and reliability. Validity relates to whether the right quantity is measured and is assessed by comparing a metric with a gold-standard metric. Reliability relates to whether measurements are repeatable and is assessed by comparing repeated measurements. The accuracy and precision with which measurements are made affect both validity and reliability. A major source of unreliability in behavioural data comes from the involvement of human observers in the measurement process. Where trade-offs are necessary, it is better to measure the right quantity somewhat unreliably than to measure the wrong quantity very reliably. Floor and ceiling effects can make measurements useless for answering a question, even if they are valid and reliable. Outlying data points should only be removed if they can be proved to be biologically impossible or to result from errors.