Madrigal Chapter 1 Flashcards
Introduction to statistics (26 cards)
Two main reasons statistics are integral to anthropology
- Allows us to approach our subject of study in a manner which lets us test hypotheses about the subject.
- If we quantify results, we can compare them. Need for comparing is because scientific results should be replicable
Fact
Verifiable truths, where the verification is not limited to human senses
Hypothesis
“an explanation of facts” according to this shawty
Can be tested and rejected by empirical evidence
Cannot be proven to be true, but can be rejected and the less rejected, the more likely it is to be correct.
Theory
A set of unified hypotheses - none which has been rejected - but if they do, the rejected hypothesis is revisited (without the theory falling apart)
Good scientists never say they have found an absolute truth! Science is a changing field
Statistical methods
Methods to test hypotheses in science
Anthropology as a comparative science
Cross-cultural comparative view of anything human is embedded - with statistics, we are able to compare results with the results of others
Unit of analysis
The factors studied in the data set - including variables and constants
Constants
Observations that are recorded on the subjects which do not vary in the sample.
Eg every subject being women living in the same village
Variables
Observations which vary from subject to subject
A singular observation in a subject
- An observation
- An individual variate
- Datum
Qualitative variables
Classify subjects according to the kind of quality of their attributes.
Often referred to as attributes, categorical or nominal variables
Qualitative variables in a coding system
Consists of
1) Mutually exclusive categories (each observation placed in one category)
2) Exhaustive categories (all observations should be categorised
Ranked or ordered variables
Observations that can be ordered from a lower to higher rank. The distance is not fixed or set.
Numeric or quantitative variables
Measure the magnitude or quantity of the subjects’ attributes
Divided into
1) Discontinuous numerical variables - no intermediate values between them - but distance between value fixed
2) Continuous numeric valuables - numeric data which allow an infinite number of values between two data points.
- Interval and ratio fall under here (interval=0 has meaning, ratio = 0 means absent)
Accurate vs precise measurement
Accurate measurement: One that is close to the true value of that which is measured
Precise: One that yields consistent results
A precise measure may not be accurate.
Experimental group vs control group
Experimental: receives a treatment
Control: remains undisturbed and serves as a comparison point
Can show if a change does or does not affect subjects
Results are dependent on random sampling of groups!
Statistical population
The entire group of individuals the researcher wants to study - often incompletely observable
denoted with N
Parameter
Measure, eg mean, that characterises a population and is denoted with greek letters. The value of population parameter is usually unknown.
Sample
Subset of the population, generally provides data for the research
denoted with n
Statistic
Measure that characterizes a sample
Descriptive statistics
Describing the sample by summarizing raw data
Include measures of
1)Central tendency
2) Dispersion
Central tendency
The value around which much of the sample is distributed
Eg mean, mode, median
Dispersion
How the sample is distributed around the central tendency value
Eg range, frequences, IQR, standard deviation
Inferential statistics
Statistical techniques which use sample data, but make inferences about the population from which the sample was drawn.
