Please enable JavaScript.
Coggle requires JavaScript to display documents.
Bi-variate Data (Line of best fit (LOBF) (Linear regression, y(hat) vs y,…
Bi-variate Data
Line of best fit (LOBF)
Linear regression
y(hat) vs y
least squares regression line of y on x
y on x predicts y based on x (and vise versa)
Further analysis
form
linear
non-linear
direction
positive
negative
strength
strong
moderate
weak
Residual Plot
Linear regression techniques to model sets of ordered pairs
residual = random for linear, pattern for non-linear
residual = y- yhat
residual = real plot - lobf plot
Row and column percentages
e.g.
Association if one
seems
to impact the other
Scatter graphs
General association/correlation seen in trend
Explanatory variable (independent) vs. Response variable (Dependent)
Reliability of prediction
Interpolation (between given points) - reliable
Extrapolation (beyond given points) - unreliable
Pearson's Correlation
Describes relationship between 2 variables
strength of relationship found in terms of (r)
closeness to linear relation (rxy)
1 = perfect positive
-1 = perfect negative
Cause and Effect
cannot assume change in 1 causes change in another (suggest relationship)
association =/= cause and effect
r^2 - The Correlation of Determination
what proportion of the total variation of y is bc of x
how much of it can be explained by a linear relation