Biometry quiz 3

pvalue

describes how likely it is that your data would have occurred by random chance

variance

∑(xi- 𝓍)^2 / N-1

Sum of squares

∑(xi- 𝓍)^2

mean

total/n

covariance

∑(xi-x)(yi-y)/N-1

purpose of normality test

asses if the dataset is normally distributed

Monotonic

: Data values are changed but value ranks are not changed

Non-monotonic

Data values are changed but value ranks are also changed

parametric test

pearson r

nonparametric tests

kendall's tau, spearman rank

Nonlinear association is called

association

Assumptions for correlation

1. Random sampling
2. Variables are interval or ratio measurements
3. Variables normally distributed, no outliers
4. Distributions have equal variances
5. Linear relationship between two variables

independent t-test

compares 2 means from independent populations (data from different groups of ants)

dependent t-test

compares two means based on related observations

t test equation

(observed mean 1- observed mean 2)-(expected mean 1-expected mean 2)/Standard error difference (mean 1-mean 2)

independent t-test

(observed mean 1- observed mean 2)/ sqrt (sample variance)

r equation

r=sqrt t^2/t^2+df

Dependent t-test

observed mean difference-expected mean difference/ standard error

dependent

y-axis

independent

x and z axes

Modeling in the research process

1. observation from data
2. generate a theory
3. generate hypothesis-identify variables
4. collect data to test the theory-measure variables
5. analyze data - graph data and fit model

How can normality be assessed

Graphically, quantitatively, statistically

Log transformation

high degree of variation, x is positive, lots of zeros.
(reduces skew)

Power exponents

reduces positive skew

Arcsine transformation

data must be between 0 and 1
(proportional data)

Non-monotonic

presence/absence (0/1)

Reciprocal transformation

1/x
reverses scores (nonmonotonic)