Biostatistics: essential concepts for the clinician

TORRES, Darlyane; NORMANDO, David

doi:10.1590/2177-6709.26.1.E21SPE1

Darlyane TORRES

Universidade Federal do Pará, Programa de Pós-Graduação em Odontologia (Belém/PA, Brazil)

https://orcid.org/0000-0002-7923-9967

David NORMANDO

Universidade Federal do Pará, Departamento de Odontologia (Belém/PA, Brazil)

https://orcid.org/0000-0002-1335-1040

davidnormando@hotmail.com

AUTHORS' CONTRIBUTION

Darlyane Torres (DT)

David Normando (DN)

Conception or design of the study:

DT, DN.

Data acquisition, analysis or interpretation:

DT, DN.

Writing the article:

DT, DN.

Critical revision of the article:

DT, DN.

Final approval of the article:

DT, DN.

Overall responsibility:

DN.

The authors report no commercial, proprietary or financial interest in the products or companies described in this article.

	NUMERIC OR QUANTITATIVE VARIABLES It is expressed in numbers		NON-NUMERIC, CATEGORICAL OR QUALITATIVE VARIABLE It is expressed in words
	Discrete	Continuous	Ordinal	Nominal
Concept	It only assumes integer values such as 0, 1, 2, 3, 4 and so on, not allowing fractional values. It is related to counts.	Assumes numeric values both integer and fractional (decimal). It is related to the measurement of quantities.	Represents two or more categories in which the data has ordering or hierarchy.	It represents two or more categories in which there is no order or hierarchy.
Examples	DMFT index; number of erupted teeth.	Cephalometrics measurements; Anterior open bite, in millimeters; Treatment time, in months.	Education level; Pain intensity (absent, low, moderate, severe); Plaque index.	Gender: female or male; Blood type: A, B, AB or O; Angle classification: I, II or III; Questions where the answers can be “yes” or “no”.

	DEPENDENT VARIABLE Called “Response Variable”	INDEPENDENT VARIABLE Also known as “Explanatory” or “Predictor”
Concept	It is the event or characteristic that you want to discover or explain. It represents a quantity whose appearance, disappearance, increase, decrease, etc. depends on how the independent variable is handled by the researcher.	It is the determining factor, condition or cause that makes it possible to predict a response, effect or consequence. It can vary during the study or be controlled, but is not affected by any other variable within the experiment.
Example	In one study, it is intended to ascertain the need for orthodontic treatment based on gender, age, education, socioeconomic level and perception of oral health. Thus, the response variable (dependent) of the study is the “need for orthodontic treatment”, while the others are explanatory (independent) variables.

	FREQUENCY DISTRIBUTION Used for both numeric (quantitative) and non-numeric (qualitative) data. Count of occurrence in the sample by frequency, rate or ratio		SUMMARY MEASURES The summary provides the distribution of data, by central tendency or by variability
	FREQUENCY	RATE OR RATIO	CENTRAL TENDENCY	VARIABILITY
C O N C E P T	Absolute frequency: Represents the number of times that each category appeared in the sample. It is the result of counting the number of sample units belonging to each category. Relative frequency: Assumes numeric values both integer and fractional (decimal). It is related to the measurement of quantities. It is the division between the absolute frequency of a category and the total frequencies observed. Percent frequency: It is the relative frequency multiplied by 100%.	Rate: It is the relative frequency multiplied by 1000, 10000 or 100000. Ratio: It is the relative frequency of one category divided by the relative frequency of another category.	Mean: Is the quotient between the sum of the data and the total number of observations (n). It should be used in quantitative data with normal distribution. Median: Represents who is in the middle of the ranked sample. It can be used in non-normal quantitative or ordinal data. When normal distribution is achieved, its value is similar to that of the mean. Mode: Represents the most often value. It can be used in any type of variable, but it is of little use in publishing scientific studies	Also known as “dispersion measures”, as they reveal how the data varies or is distributed around its midpoint. Amplitude: Is the difference between the highest and lowest value in a data set. Standard deviation (SD): Is the value that represents the symmetric average dispersion of data around the mean of a data set. It is used in quantitative data with normal distribution. Variance: Is the standard deviation value raised to the square. Variation coefficient: Is the relative dispersion of the data, represented by the ratio between the standard deviation and the mean, multiplied by 100. Percentile: Percentiles are the 99 values that separate a series into 100 equal parts. Quartile: Quartile are the values of a series that divide it into four equal parts. First quartile (Q1) includes the first 25% of the data, second quartile or median (Q2) includes the first 50% of the data, third quartile includes the first 75% of the data (Q3). Interquartile deviation: Difference between the Q3 (P75) and Q1 (P25), which is not influenced by extreme values. Should be used when non-normal data are being evaluated.

[1] davidnormando@hotmail.com

Brasil

Brasil

Biostatistics: essential concepts for the clinician

ABSTRACT

Introduction:

Objective:

Conclusion: