Crosswalk Data Analysis - Southern Utah University

Download Report

Transcript Crosswalk Data Analysis - Southern Utah University

Crosswalk
Data
Analysis
Add the names of all
the team members to this
first slide
Lynn White’s Stats Class
Spring 2011
Our Study
State the general purpose of your study
 The Main Purpose: We wanted to investigate flag
use at a pedestrian crosswalk and the variables
that determine whether a flag is used or not.
Procedures
Give a brief description of the procedures
And/or the survey you used
 Data was collected…
 At the crosswalk across the street from Lins
 In pairs
 One individual on each side of the street
 The individual on the side opposite the way individuals
crossed administered the survey
 On both weekdays and weekends
 During daylight hours
 In clear weather
IV’s and DV’s Manipulated or Recorded
Flag/No Flag
Solo/Group Crossing
Group Size
Age
Gender
Resident
Highest Level of Education
Traffic
If children under 18 were present
Our Sample
Use descriptive stats to describe your
Sample (sample size, # males/females,
Mean age and Standard deviation… other
Important characteristics to define the
Sample). Do not report any SEMs here.
 n= 45
 Mean Age: 28.23 years old, s = 2.33
 Men = 21 women = 24
What to put on each slide for the statistical tests
State the research question you are using statistics to answer
be careful with word choice (relationship vs. difference)
When using a t-test or ANOVA: state the IV(s), the levels (separately for
each IV - if you have 2 IVs), the DV being analyzed, and the scale of
measurement. Make sure you state the FULL name of the test.
When using correlation or chi square, state what the two variables are and
their level of measurement (hint: for correlation, both need to be
Interval or ratio; for chi square, both need to be nominal or ordinal). For chi square, state
the FULL name of the test.
Put in a graph (histogram or polygon) with SEMs to show differences between means
Put in a scatterplot to show a correlation ; Put in a table to show chi-square
For each statistical test, you must write out the results in correct APA format. Then state
the finding in plain simple English (there are samples in my stats book). Make sure you tell
me if the results are significant or not significant.
AESTHETICS COUNT! Make your presentation look DYNAMITE!
Does Age Differ by Residency Status?
 Independent t-test
IV = residency (CC resident vs. non-CC resident),
DV = age (ratio)
 Not significant t(44) = 1.86, p > .05
 Mean age did not differ between residents vs. nonresidents of Cedar City
Mean age (years)
29.5
29
28.5
28
27.5
27
Flag
resident
No Flag
Non-resident
Does Flag use Depend on Traffic Congestion?
 chi square contingency test: variable 1= flag use
(nominal), variable 2 = traffic (ordinal)
 Not significant χ² (5, N= 45) = 2.20, p > .05.
 Flag usage does not depend on whether the traffic was light,
moderate or heavy.
Flag usage as a function of traffic
light
moderate
heavy
No flag
55%
49%
44%
Flag
45%
51%
56%
Note: % are percent of sample within each traffic category
Is years of education correlated with the time it
takes to cross the intersection?
 Pearson correlation: variable 1 = years of education
(ratio), variable 2 = time to cross the intersection (ratio)
 Significant, r= .79, p= .000
 As the number of years in school increased, time to cross
increased.
35
30
Time (sec)
25
20
15
10
5
0
0
5
10
15
Years of education
20
25
Does Age Differ by Gender and Flag Use?
 2 Way independent ANOVA
mean age





40
35
30
25
20
15
10
5
0
IV: gender (men vs. women, between subjects)
IV: flag usage (no flag vs. flag, between subjects)
DV = age (ratio)
Neither main effect was significant at p > .05
The interaction was significant F(1, 41)= 5.49, p=.002
men
women
The mean age of women was
the same for flag carriers and
non-carriers.
The mean age of men who
carried flags was significantly
higher than the mean age of
men who did not carry a flag.
no flag
flag
Conclusions
Mean age did not differ by residency status
Flag use did not depend on traffic congestion
Years of education was positively correlated with
time to cross the intersection
For women: mean age did not differ between
Carriers and non-carriers
For mean: flag carriers were significantly older
Than non-carriers