Transcript Results & Data Analysis

RESULTS & DATA ANALYSIS
Descriptive Statistics

Descriptive (describe)
 Frequencies
 Percents
 Measures
 mean
 median
 mode
of Central Tendency
Measures of Spread

Range
 Nominal
data = number of responses in each
category (A’s = 4, B’s = 7, C’s = 21, D’s = 1, F’s=2)
 Other data = difference between responses for
the greatest and least numeric values (Age of
oldest is 104 and youngest is 18. Range =86 years)

Tertiles, Quartiles, Quintiles
 Interquartile Range (IQR)
 Range for the 25th to 75th
percentiles
which captures the middle 50%
Measures of Spread
 Standard
Deviation – describes, on
average, how much individual values differ
from the mean
 Standard
Scores or Z-scores – describes
how many SD’s away from the sample
mean an individual score or response is
Normal Distribution
68% of individuals
95% of individuals
>99% of individuals
Sample
Mean
-3 SD
-2 SD
-1 SD
+1
SD
+2
SD
+3
SD
Inferential Statistics
(Comparative)

Based on Probability

Tests of significance: are observed differences
real differences or simply the result of chance
 t-test:
test difference between means of two groups
(t statistic)
 ANOVA
(Analysis of Variance): test difference among
three of more independent groups (F statistic)
Interpreting p-values or
probability value


Used to decide whether the results observed are
likely to reflect real differences between groups
The standard is to use α = 0.05 or 5% (1 in20)
 t = 1.11, p =0.042
 t = 1.02, p = 0.051


Probability that results observed have occurred by
chance alone
α = probability of Type I error (finding statistical
significance when in reality there is none)
Significance

Statistical

Practical

Clinical
Parametric and Non-Parametric

Parametric = DV is some measured quantity
(ratio or interval) so it makes sense to calculate
means and SD
 You
can draw bell curves through the data defined
by two parameters…the mean and SD….hence
parametric (beside, near)

Non-Parametric = DV is count or rankings so
means and SD have no meaning (e.g. The
average religion of Americans is 2.67)
Parametric Tests

T-tests

ANOVA (analysis of variance)

Linear Regression or Multiple Linear Regression

ANCOVA (analysis of covariance) – combines
regression and analysis of variance
Non-Parametric Tests

Chi-square

Logistic Regression

Log-linear analysis