E0410 Fundamentals of Statistics for Scientific Data Using R
by Daria Sapunova, PhD student, RECETOX
daria.sapunova@recetox.muni.cz
Bohunice, D29, room 123

Repetition

Mentimeter.com

Association between multiple variables:
quantification
We have: multiple numerical and categorical variables
Height Gender Smoker
Weight
Salary Education Urbanization
Productivity
Teenage girl measures height standing near marks on wall wanting to become tall Man paying online
and receiving cashback to wallet Losing weight. Female nutritionist cartoon character. Slimming,
weightloss, dieting. Counting calories. Overweight man with hamburger. No Smoking Sign City skyline
concept illustration Online certification illustration

Association between multiple variables:
quantification


Check mark and cross symbols in flat styles Check mark and cross symbols in flat styles
     Weight (kg)
     Height (cm)
One-unit shift
How much will change?
?
man character standing icon man character standing icon
+1 cm
0.87 kg
     Gender
Association between multiple variables:
quantification
Male
Female
     Gender
     Height (cm)
One-unit shift
     Weight (kg)
?
How much will change?
Male and female icon. Man and woman gender symbols. User avatars. Gentleman and lady. Male and
female icon. Man and woman gender symbols. User avatars. Gentleman and lady.
2.14 kg

Multiple linear regression
The basic format of a multiple  linear regression is:
Y = α + β1*X1 + β2*X2
Where:
Y = outcome (i.e. dependent) variable.
X1 = first predictor (i.e. independent) variable.
X2 = second predictor (i.e. independent) variable.
α = intercept (average Y when X1=X2=0). Note: α is unit specific.
β1 = slope of the line (change in Y for a 1 unit increase in X1 when X2 is held constant). Note: β1
is unit specific.
β2 = slope of the line (change in Y for a 1 unit increase in X2 when X1 is held constant). Note: β2
is unit specific.
https://www.coursera.org/learn/linear-regression-r-public-health;
https://towardsdatascience.com/graphs-and-ml-multiple-linear-regression-c6920a1f2e70

Association between multiple variables:
quantification
We have: multiple numerical and categorical variables
Fruit consumption; veggie consumption; fish consumption; meat consumption
Mercury level
Be aware: collinearity as an assumption in case of multiple numerical variables!
Pollution toxic environment damage and global contamination flat isolated Bright vector
illustration of colorful juicy orange isolated, organic citrus fruit vegetarian food menu Vintage
fish illustration Set of pieces of meat of varying degrees of roasting




Practical part