* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Lecture6 - University of Idaho
Survey
Document related concepts
Transcript
PSYC512: Research Methods Lecture 6 Brian P. Dyre University of Idaho PSYC512: Research Methods Lecture 5 Outline Questions about material covered in Lecture 5 Scientific Method: Proof and disproof & Strong Inference Operational definitions Issues in Measurement Choosing Measures Scales of Measurement Variables: Reliability and validity sampling PSYC512: Research Methods Choosing Measures Research tradition e.g., operant conditioning—lever pressing e.g., cognition—accuracy and reaction time e.g., sensation and perception—discrimination accuracy e.g., personality—surveys, inventories (self-reports) Theory e.g., the psychophysical postulate – discrimination accuracy e.g., Serial vs. parallel processes in visual search – RT Availability of new techniques Availability of equipment PSYC512: Research Methods Features of Measures: Scale of Measurement (Stevens, 1946) Four types: nominal, ordinal, interval, and ratio Nominal scales set of unique cases, types, or categories with NO ORDER Only non-parametric operations are valid: counting frequencies, modes, chi-square, point-biserial correlation Ordinal scales different categories that can be ranked along a continuum more or less, but not how much more or less Only non-parametric operations are valid : counting frequencies, modes, medians, chi-square, rank-order correlation PSYC512: Research Methods Features of Measures: Scale of Measurement (Stevens, 1946) Interval intervals of the scale are equal in magnitude Necessary but not sufficient condition for parametric statistical tests valid operations: all mathematical operations, means, standard deviations, etc. may be calculated If other distributional assumptions are met: linear and non-linear regression, t-tests, ANOVA are also valid no fundamental zero—no ratio statements allowed Ratio Like interval but also has a fundamental zero point—allows ratio statements such as “A is twice as much as B” Generally interval or ratio scales should be used if possible More powerful and flexible statistical tests More precision in evaluating quantitative hypotheses PSYC512: Research Methods Features of Measures: Sensitivity Sensitivity: measure must show changes in response to changes in the independent variable Range effects Ceiling effects: variable reaches its highest possible value and gets truncated (test is too easy) Floor effects: variable reaches its lowest possible value and gets truncated (test is too hard) PSYC512: Research Methods Features of Measures: Reliability the ability of a measure to produce consistent results when repeated measurements are taken under identical conditions Types: precision: physical measurement (1/noise) margin of error: sampling in surveys interrater reliability: observers viewing the same behavior Test-retest, parallel forms and split-half reliabilities: psychological tests PSYC512: Research Methods Other Features of Measures Accuracy does a measure produce results that agree with a known standard? Accuracy vs. Precision Validity Measurement validity: the extent to which your measure indeed measures what it is intended to measure Types: Face validity, Content validity, Criterionrelated validity (concurrent vs. predictive), Construct validity Relationship between reliability and validity PSYC512: Research Methods Probability and Statistics Why are probability and statistics important? Used to assess variability in data Treatment Variance Variability due to different levels of independent variable Good variance that we want to maximize Error Variance Variability in data due to factors other than the treatment Bad variance that we want to minimize Probability and Statistics are simply tools used to assess and compare these sources of variability PSYC512: Research Methods Visualizing Variability: Distributions of Frequency and the Histogram Histograms: used to represent frequencies of data in different classes or categories Bin 6 Frequency 4 2 0 0 1 2 3 4 5 6 7 8 9 10 Grade PSYC512: Research Methods Frequency 0 0 1 0 2 0 3 0 4 3 5 1 6 6 7 4 8 2 9 6 10 1 Displaying Histograms: Stem and Leaf Plots Stem and Leaf plots are used to display histograms graphically (on their side) using only typed characters Stem Leaf (hypothetical histogram for IQ) 6 7 8 9 10 11 12 13 78 35668 012234445555667777889 00011233333334445566667889999 01112233334444445566677777888899 0001122233444566777899 0012569 02 PSYC512: Research Methods Distributions of Probability Density Similar to frequency histogram except y-axis now represents probability density (mass) rather than frequency Probability density = Frequency/N 0.5 Probability Density 0.4 0.3 0.2 0.1 0.0 0 1 2 3 4 5 6 Grade PSYC512: Research Methods 7 8 9 10 Some Types of Distributions Gamma 0.14 0.18 0.12 0.16 Probability Density Probability Density Normal 0.10 0.08 0.06 0.04 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.02 0.00 0.00 0 5 10 15 20 25 Data 0 5 10 Data PSYC512: Research Methods 15 20 25 Measures of the Center of a Distribution Measures of center represent the general magnitude of scores in a distribution Mode: most frequent score Median: the middle score of an ordered distribution Mean (average): where X is the data and N is the total number of observations X N PSYC512: Research Methods Measures of the Spread of a Distribution Measures of spread are used to assess the consistency of scores in a distribution Range = max score – min score Interquartile range = score(Q3) – score(Q1) Variance (s2) and standard deviation (s) where X is the data, 2 m is the mean of the data, X and N is the total number s2 N of observations PSYC512: Research Methods More on Variance Standard Deviation (s) = sqrt(variance) where X is the data, 2 X m is the mean of the data, s and N is the total number N of observations Why N instead of N-1? Populations vs. Samples Remembering how to compute variance “the mean of the squares – square of the means” s 2 X N 2 X N 2 PSYC512: Research Methods Describing Distributions Parametrically: Statistical Moments Any distribution based on interval or ratio data can be summarized by its statistical moments First Moment: Mean—location of distribution on x-axis Second Moment: Variance—dispersion of distribution Third Moment: Skewness—symmetry of distribution Fourth Moment: Kurtosis—degree of “peakedness” PSYC512: Research Methods Estimators Sample statistics estimate population parameters Mean: M or X vs. M Variance: s2 vs. s2 Properties of Estimators Sufficiency: uses all information in sample (mean and variance are sufficient, mode and range are not) Unbiasedness: expected value approaches real value with increased sampling Efficiency: tightness of cluster of sample statistics relative to the population parameter Resistance: influence of outliers on sample statistic PSYC512: Research Methods Next Time… Topic: descriptive statistics, variables, sampling, and more on hypothesis testing Be sure to: Read the assigned readings (Howell chapters 3-4) Continue searching and reading the scientific literature for your proposal PSYC512: Research Methods