Module 14 Research methods Flashcards
Statistics
Branch of math that seals with analysis of numerical data
Descriptive statistics
describe, summarize, or make sense of a particular set of data
Data set
collection of related sets of info
Graphs
pictorial representations of data in 2D space
Bar graph
vertical bars that represent data
Categorical variable
Bars do not touch
Histogram
graphic representation of frequency distribution
Quan variable
Bars touch
Line graph
Representation of linear data
Scatter plot
way to visualize relationship between two quantitative variables
Is there a relationship?
Is it linear or curvilinear?
Is linear relationship positive or negative?
Is linear relationship weak or strong?
Measure of central tendency
numerical value most typical of the values of a quan variable
Mode
most frequently occurring number of a set
Median
Middle point of a set of numbers
for even quantity, find average of middle two numbers
Mean
arithmetic average
Distribution of data
Normal distribution - bell curve
Negatively skewed - tail to the left (toward negative numbers)
Positively skewed - tail to the right (toward positive numbers)
Inferential statistics
Use laws of of probability to make inferences about populations based on samples
Estimation
rough calculation of the value, number, or quantity of something
Point estimation
Using a single number as the estimate of the population parameter
Interval estimation
Includes a range of numbers thought to contain the population parameter
Significance tests
Researchers report statistical significance to enhance credibility
T-test
used for independent samples or correlation coefficients
Analysis of variance (ANOVA)
Analyze group means of a sample
Chi-square test
For contingency tables
Used to test relationships between categorical variables
Hypothesis testing
How well the sample data support the particular hypothesis (null) and when the null can be rejected
Null hypothesis
predicts no difference or relationship in the population
Alternative hypothesis
the population parameter predicts a value other than the null
