Biostatistics and General concepts/jeopardy

Measurement; Quantitative; Qualitative

_____ is the process of collecting and recording observations about variables of interest, and has ____ and _____ data

summarization; hypothesis testing/interpretation

Descriptive statistics is the ______ of data, while inferential statistics is the ______ 

Numerical (quantity) and categorical (quality)

The two types of VARIABLES

Continuous; Discrete

Numerical variables can be broken down into ____ and _____ variables

• any value within a range

• can be interval data or ratio data

• ex. weight, height, temperature, BP

Continuous numerical variables include: 

• only taking specific, finite values

• usually whole numbers

• Ex. # hospitalizations (cant be 1.5), # days without medication

• THINK WHOLE NUMBERS!!

Discrete numerical variables include:

Nominal; Ordinal

Categorical variables can be broken down into ____ and _____ variables

Variables that cannot be placed in a meaningful order

• Ex. Sex, race, type of pharmacy, etc.

Nominal variables include:

• being classified into a logical order list

• Ex. Likert-type scales

• Pain scales

Ordinal variables include:

independent variable

This variable is the variable hypothesized to explain an observed clinical phenomenon; (explain or predicts the values of dependent variable)

Dependent variable

This variable is the observed clinical phenomenon (THE OUTCOME)

Control variables

Other explanatory variables that are presumed to influence the dependent variable

Mean, Median, Mode

The three measures of CENTRAL TENDENCY (Central - think mmm)

Mean

The ____ is sensitive to outliers

Median

The ____ is robust to outliers

Range, IQR, Standard Deviation, and Standard Error

The measures of dispersion are

difference between the 75th percentile and the 25th percentile of a dataset; represents the range of the middle 50% of the data (measure of dispersion not affected by extreme values).

IQR is the ___ 

the dispersion of individual data around the SAMPLE mean

Standard Deviation measures:

variability of sample means from DIFFERENT SAMPLES, and is a representation of the POPULATION mean. Larger SE means more uncertainty and less precision in a samples estimate of a population parameter that is being measured. 

Standard Error measures:

Frequency Table

The most simple form of data visualization is the:

Histograms

____ are used to display frequency distributions for continuous variables (I.e BP, temperature, etc.)

Scatter plots

____ depict the relationships between two continuous variables. USED FOR CORRELATION ANALYSIS

Box and whisker plot

An IQR can be visualized by a ____ which is used to visualize range/spread of data. The Box is the 25th and 75th percentile, while the whiskers are the minimum/max values 

bar/Graph chart

A _____ is used to present discrete categorical data

The distribution of all the possible values of a variable and their relative frequency of occurrence (Ex. observed lengths of stay for patients in a hospital)

Empirical distribution is known as:

The mathematical representation of how values of a random variable are expected to behave (ex. how theyre spread across a range of values) 

Statistical distribution is known as:

Normal Distribution

The bell curve is known as ______ and it  includes all measures of central tendency being equal

z distribution; flatter with thicker tails

For t distribution in regard to normal distribution, as sample size increases, the T distribution approaches ______, and as it decreases, it gets ________ 

Degrees of freedom (df)

The difference between t and z distributions depends on the

population standard deviation is unknown

t distribution is used when the __________ or it cannot be reliably estimated (like with a small sample size). it is also more common

Binomial Distribution 

______ is described in relation to the probability of the outcome
happening (p) and the outcome not happening (q or 1-p) and the number of trials where the outcome might occur 9works with categorical variables)

1. Chi-square

2. F distribution

3. Poisson distribution

4. Gamma distribution

4 other types of distributions

In statistical theory, a SAMPLE is a subset or portion of the population (all observations of interest) and sample size is the number of sampled observations

What is the central limit theorem?

1. The mean of all sample means will equal population mean

2. The SD of sampled means is equal to the standard error of the mean

3. As sample size increases, the distribution of the sample means approaches a normal distribution 

3 main tenets of the central limit theorem

Statistical inference

_____ is the process of analyzing data from a sample to infer the true values or true effects in the population. The main thing that can allow for this to happen is a random sample. 

1. statistical estimation (estimates of population parameters are calculated from sample statistics)

2. Hypothesis testing (Null and alternative)

The two types of complementary approaches to statistical inference

Null

In hypothesis testing, the testing is only for the ____ hypothesis

The error of rejecting the null hypothesis when the null is true (False positive) - i.e finding a difference where there is no difference

Type 1 error

5%; 95%

The type 1 error is traditionally set at ____ and would correlate with a confidence interval of _____ (1-alpha)

Error of failing to reject the null hypothesis when the null is true (False Negative) - i.e not finding a difference when there  IS one

Type II error

10-20%; 80-90%

The type II error is traditionally set at ______ which translate to power levels of ____ because power = (1-beta)

• provide info about precision of estimates (p-values alone do not)

• Narrower intervals mean mor precise estimates

• Useful for comparing effects

• Can serve as a method for hypothesis testing while adding info beyond what p-values provide

The importance of confidence intervals in statistical inference

both positive and negative directions; in only one direction

Two-tailed tests look _____ and directional tests look _____

• tests of difference (see if A and B are different)

• tests of equivalence (see if A and B are practically equivalent)

Non-directional tests include

• Test of superiority (see if A is greater than B)

• test of non-inferiority (see if A is more more than a certain amount lower than B)

Directional tests include

P-value

____ is the probability of finding the results of the study, assuming the null hypothesis is true - Or the STRENGTH of the evidence against the null hypothesis (not strength or size of findings).

it is not the probability of whether null or alternative hypothesis is true, but rather the strength of the evidence.  

The null hypothesis is NOT rejected (cannot prove/maybe there is not a difference)

If a confidence interval includes the null value (typically 0), then ______

The null hypothesis IS rejected at chosen significance level

If a confidence interval DOES NOT include the null value, then ______

whether the null hypothesis is rejected based on statistical analysis (Stats say there is a difference)

Statistical significance is

the practical importance or impact of findings on practice

Clinical significance is

Number needed to treat (NNT); Number needed to harm (NNH)

Clinical significance can be quantified by _____ and _____

underpowered

Smaller than necessary studies (small sample size) run the risk of generating invalid scientific knowledge while exposing subjects to risks and burdens, and can be classified occasionally as ______

increases = more accurate parameter estimates = greater precision

As sample size and effect size increases, power ____

sample size

If you want a small type 1 error, you need a larger ____