Research
Statistics
Using
JASP
ENGR. ELMAR C.
FRANCISCO, MSEE
Introduction to JASP
JASP - Jeffrey’s Amazing Statistics Program
in recognition of the pioneer of Bayesian inference Sir Harold Jeffreys
This is a free multi-platform open-source statistics package, developed and
continually updated by a group of researchers at the University of
Amsterdam.
They aimed to develop a free, open-source program that includes both
standard and more advanced statistical techniques with a major emphasis on
providing a simple intuitive user interface.
Reference Manual:
[Link]
[Link]
Coverage
This training includes the following:
JASP Basic Navigation
Descriptive Statistics
Plots and Tables
Hypothesis Testing and Basic Interpretation
Limitations
Due to time constraints, this training shall not include
Formula and Manual Computation (Hence we use
JASP)
Complete Discussion of the Interpreted Data
Structured Research Writing
Advanced Functionalities (Bayesian, Factor Analysis,
Linear Regression)
However, trainings related to these topics maybe
provided by TUP-IRTC for free!
JASP
Navigation
The main menu can be
accessed by clicking on the
top-left icon.
JASP
Navigation
JASP has a streamlined interface to
switch between the spreadsheet,
analysis and results views.
Data Types
Nominal - These are the set of values that don’t possess a natural ordering (Sex/Gender, Color of
smartphone, School, Nationality)
Ordinal - natural ordering while maintaining their class of values ( Letter Grading System, Letter Sizes,
Socioeconomic Class)
Scale – quantitative data
Ref: [Link]
JASP
Navigation
JASP Navigation
Click the icon at the top row of the data to change data type.
JASP Navigation
Click top row to change labels (Don’t forget to press ENTER button after typing the new label).
JASP
Navigation
Try editing the data by clicking
any data cell.
JASP Navigation
This will generate a data file (saved
in excel as .csv). Try changing the
value of a cell and save!
Descriptive
Statistics
Aims to describe a chunk of raw data using summary statistics, graphs,
and tables ([Link])
Measures of Central Tendency
Mean – average
Mode – most frequent
Median – middle value
Standard Deviation - quantify the amount of dispersion of data values
around the mean
(Low value – dikit sa mean, High Value – kalat kalat, malayo sa mean)
Descriptive Statistics
Case 1: extra
Descriptive Pill A Pill B
Statistics Valid 10 10
Missing 0 0
Find the mean, median, mode, Mean 0.750 2.330
minimum value, maximum Median 0.350 1.750
value, and standard deviation
of the Sleep Study (extra Mode ᵃ -1.600 -0.100
means the number of added Std. Deviation 1.789 2.002
hours of sleep)
Minimum -1.600 -0.100
Maximum 3.700 5.500
ᵃ More than one mode exists, only the first is reported
Descriptive
Statistics
Quartiles – dividing the
arranged data into four groups
Deciles - dividing the arranged
data into ten groups
Percentiles – into 100 groups
(for large samples)
Descriptive
Statistics
Skewness describes the shift of
the distribution away from a
normal distribution.
Negative skewness shows that
the mode moves to the right
resulting in a dominant left tail.
Positive skewness shows that the
mode moves to the left resulting
in a dominant right tail
Case 2:
Descriptive
Statistics
Kurtosis describes how heavy or
light the tails are.
Positive kurtosis results in an
increase in the “pointiness” of
the distribution with heavy
(longer) tails.
Negative kurtosis exhibit a much
more uniform or flatter
distribution with light (shorter)
tails.
Descriptive Statistics
Case 2: Pill A
extra
Pill B
Descriptive
Statistics Mean 0.750 2.330
Median 0.350 1.750
Mode ᵃ -1.600 -0.100
Add quartiles, skewness, and
kurtosis Std. Deviation 1.789 2.002
Skewness 0.581 0.386
Kurtosis -0.630 -1.421
Minimum -1.600 -0.100
Maximum 3.700 5.500
ᵃ More than one mode exists, only the first is reported
Case 2: Descriptive You can manipulate your data analysis such as
changing the variables/split settings using the arrow
Statistics keys in the main menu.
Data
Visualization
Box plot shows the ff:
Minimum value
Maximum Value
Q1 to Q3 (inside the box, IQR)
Median – line inside the box
Data Visualization
Data Visualization Distribution plots
Case 3:
Frequencies
Open Dancing Cats and Dogs
under 5. Frequencies
Provide a frequency table and
pie chart for each variable
Case 3:
Frequencies
Open Dancing Cats and Dogs
under 5. Frequencies
Provide a frequency table and
pie chart for each variable
makes the use of various analytical tools to draw inferences
Inferential about the population data from sample data (Cuemath)
Statistics Correlation, T-tests, ANOVA, etc.
Open
Correlation Analysis
Correlation is a statistical technique that can be used to determine if, and how strongly, pairs of
variables are associated.
Correlation is only appropriate for quantifiable data in which numbers are meaningful, such as
continuous or ordinal data.
Pearson’s correlation
coefficient (or "r"). It ranges from -1.0 to +1.0. ( >0 direct correlation, <0 inverse correlation)
Case 4: Correlation
Open College Success (4. Regression)
Using p value of 0.05 (5% margin of error)
Provide Descriptive Statistics and boxplots for males and females (Split: Sex)
Check correlation between
A. high school grades in Mathematics (hsm) to their SAT scores in Mathematics (satm)
B. high school grades in English (hse) to their SAT scores in Verbal knowledge (satv)
C. high school grades in Science (hss) to their college GPA (gpa)
Case 4: Correlation
Case 4: Correlation
Case 4: Correlation
Case 4: Correlation
P value < 0.05 (significant)
P value > 0.06 (insignificant)
A. With a Pearson’s r of 0.454 ( p < .001 ), there is
a significant correlation between the Mathematics
grades of the students in high school to their
Mathematics scores in SAT
Case 4: Correlation
B. With a Pearson’s r of 0.244 ( p < .001 ), there is a correlation between the English grades of the
students in high school to their verbal knowledge scores in SAT
Case 4: Correlation
C. With a Pearson’s r of 0.244 ( p < .001 ), there is a correlation between the Science grades of the
students in high school to their college GPA.
Case 4: Correlation
Ref: M. M. Mukaka, “A Guide to Appropriate Use of Correlation Coefficient in Medical
Research,” MMJ, Vol. 24, No. 3, 2012, pp. 69-71. [Citation Time(s):2]
Cross
Pearson's Correlations
Variable gpa hss hsm hse satm satv
correlations 1. gpa Pearson's r
p-value
—
—
2. hss Pearson's r 0.329 —
p-value < .001 —
3. hsm Pearson's r 0.436 0.576 —
p-value < .001 < .001 —
4. hse Pearson's r 0.289 0.579 0.447 —
p-value < .001 < .001 < .001 —
5. satm Pearson's r 0.252 0.240 0.454 0.108 —
p-value < .001 < .001 < .001 0.106 —
6. satv Pearson's r 0.114 0.262 0.221 0.244 0.464 —
p-value 0.087 < .001 < .001 < .001 < .001 —
To test significant difference among two
groups of data
Case 4: 1. Independent Samples – different
t-tests groups (Group A or B)
(Independent
Samples)
2. Paired-Samples – same group for two
different conditions usually with one
group in the effectiveness of intervention
(pretest/posttest),
Case 5:
t-tests
Open Directed Reading
Activities (2. t-tests) and
identify if there is a significant
difference between the control
group and treatment group in
terms of reading power test
performance (drp)
Case 4: t-tests
With a p value of 0.029 (p value is lower than 0.05), there is a significant difference between the
directed reading power test performance between the treatment group and the control group. (two-
tailed t-test)
Case 5:
t-tests
With a p value of 0.014 (lower
than 0.05), the reading power
test performance of the
treatment group is significantly
higher than the reading power
test performance of the control
group. (one t-test)
Case 6: t-
tests (paired
samples)
Open Moon and Aggression (2.
t-tests)
Check if the disruptive
behaviors of people with
Dementia during days with full
moon is significantly higher
during days without it.
Case 6: t-
tests (paired
samples)
With a p value of < .001, the
disruptive behaviors of people
with Dementia during days with
full moon is significantly higher
than during days without it.
With a p value of < .001, the data
suggests that the individuals with
Dementia has more disruptive
behavior during full moon.
Case 7:
Analysis of
Variance -
ANOVA
For three or more groups, use
ANOVA.
Open Beer Goggles
(2. ANOVA)
Analysis of Variance - ANOVA
With a p value of 0.034, there is a significant difference in the perceived attractiveness among groups of
individuals who drank non-alcoholic, average strength, and strong alcohol.
Analysis of Variance - ANOVA
If there is a significant difference among groups, simple ANOVA is not enough. There must be post-
hoc analysis to cross-compare one group to another.
Case 8: Analysis of 0 – non alcoholic
Variance - ANOVA 1 – average strength alcohol
2 – strong alcohol
Case 8: Analysis of Variance -
ANOVA
With a p value of 0.379, there is no significant difference
between the perceived attractiveness of the respondents who
drank non-alcoholic to those who drank average strength of
alcohol.
With a p value of 0.034., there is a significant difference
between the perceived attractiveness of the respondents who
drank non-alcoholic to those who drank strong alcohol.
With a p value of 0.441, there is no significant difference
between the perceived attractiveness of the respondents who
drank non-alcoholic to those who drank average strength of
alcohol.
END OF
LECTURE ON
JASP! XIE XIE !
Next CHATGPT !
