Measurements in population

Question 1

Absolute risk

Answer 1

The probability of an event in a population under study [contrasted with the RELATIVE RISK]. Synonymous with INCIDENCE.

Question 2

Absolute risk reduction

Answer 2

The amount, sometimes expressed as a percentage, by which the risk of a disease is reduced by elimination or control of a particular exposure. It is possible from this to estimate the number of people spared the consequences of an exposure.

Question 3

Adverse reaction

Answer 3

Syn: adverse outcome. An unwanted or undesirable consequence of a diagnostic, preventive, or therapeutic regimen or procedure, e.g. illness caused by immunization against childhood infections.

Question 4

Age specific (incidence) rate

Answer 4

An incidence rate for a specified age group. The numerator and denominator refer to the same age group.

Example:
Number of deaths among residents
Age-specific death rate = age 25-34 in an area in a year x 100,000/Average (e.g. midyear) population age 25-34 in the area in that year

Question 5

Age specific mortality rate

Answer 5

A specific type of incidence rate in which the numerator is the number of deaths observed in a specified age group.

Question 6

Age standardisation

Answer 6

Syn: age adjustment. Mathematical procedures for adjusting rates in two or more populations with differing age structures so as to adjust for rate differences among them that are or could be attributable to the effect of age differences, rather than other factors.. See also STANDARDIZATION.

Question 7

Algorithm

Answer 7

A systematic process consisting of an ordered sequence of stages in which each stage is determined by the outcome of the stage that preceded it.

Question 8

Analysis of variance

Answer 8

Syn: ANALYSIS OF VARIANCE. A set of statistical methods for assessing the contribution of independent variables to the mean value of a continuous dependent variable.

Question 9

Arithmetic mean

Answer 9

A measure of central tendency, computed by adding all the values in the set and dividing by the number of values in the set. See also AVERAGE.

Question 10

Attack rate

Answer 10

The CUMULATIVE INCIDENCE rate of a condition in a particular population or group of people such as patients admitted to a hospital.

Question 11

Attributable risk

Answer 11

The risk of disease or death in the exposed group minus the risk of disease or death in the unexposed group.

Question 12

Binary variable

Answer 12

A variable having only two possible values, e.g., on or off, 0 or 1, male or female. Also known as a discontinuous variable. Cf. CONTINUOUS VARIABLE.

Question 13

Birth cohort

Answer 13

The component of the population born during a specified period, generally a year, sometimes a 5- or 10-year period

Question 14

Birth rate

Answer 14

The number of live births in a defined population, in a (usually) calendar year divided by the midyear population, with the customary multiplier of 1,000 to produce a whole number rather than a decimal or faction.

Question 15

Censoring

Answer 15

A term used in biostatistics to describe loss of cases to follow-up.

Question 16

Census

Answer 16

A periodic enumeration of the population, primarily intended to collect information to identify eligible voters, tax payers, and sometimes to identify people eligible for military service, and incidentally many other useful facts, e.g., about housing conditions. Censuses are of two broad types – de facto, in which people are enumerated where they are on a given day or night and de jure, when they are enumerated related to where they habitually live. The UK censuses, carried out every 10 years are de facto, thus, for university students, for example, if the census day or night falls in university term they will contribute to the numbers of people in the relevant university town or city. If the day or night falls during university vacation, they are enumerated in relation to where they are at that time – most will be elsewhere.

Question 17

Central tendency

Answer 17

The grouping of numerical data about a value at or near the midpoint of the overall distribution. Three common MEASURES OF CENTRAL TENDENCY are mean, median, and mode

Question 18

Chi-square (χ2) test

Answer 18

A common statistical test for detecting whether two or more populations differ from one another with respect to the proportions of people in those populations that have a given characteristic. Cf. STUDENT’S T TEST.

Question 19

Cluster sampling

Answer 19

A sampling method is which each unit selected is a group of people (e.g. a family, a household, a school or class within a school).

Question 20

Competing risk

Answer 20

An event that removes a study subject from experiencing the outcome that is under investigation. For instance, in a study of smoking in relation to lung cancer, death of coronary heart disease or in a road traffic accident is a competing risk.

Question 21

Confidence interval

Answer 21

The range within which the true value of a variable such as a mean, proportion, or rate lies, with a probability that can be calculated using statistical methods. The end points of the confidence interval are the confidence limits.

Question 22

Confidence limits

Answer 22

The upper and lower boundaries of the confidence interval.

Question 23

Contingency table

Answer 23

A table that arranges sets of data in columns and rows so that relationships between the sets can be studied by means of suitable statistical tests, e.g., the CHI-SQUARE TEST.

Question 24

Continuous variable

Answer 24

Data (variables) with a potentially infinite number of possible values along a continuum. Data representing a continuous variable include height and weight, BMI, age, blood pressure etc. Cf. BINARY VARIABLE.

Question 25

Control group, controls

Answer 25

Subject with whom comparison is made in a case control study, randomized controlled trial, or other variety of epidemiological study.

Question 26

Convenience sample

Answer 26

A sample, usually of people, that has been collected by expedient means, such as that they happen to be available for study, not by using a random sampling method.

Question 27

Correlation coefficient

Answer 27

A statistical measure of the degree of relationship between two variables. It is represented by the symbol r and in practice it usually refers to a linear relationship that in theory can vary from +1 when there is perfect correlation to -1 when the two variables are perfectly inversely related.

Question 28

Crude death rate

Answer 28

Syn: crude mortality rate. The frequency of death in a specified population during a specified period. It is calculated by dividing the number of deaths in the period by the estimated number of people at risk of dying during this period, usually a calendar year. It is called the “crude” death rate because no adjustment is made to allow for age composition of the population or for other conditions or circumstances. Thus, comparisons of crude death rates in different populations have limited value and must be interpreted with caution.

Question 29

Cumulative incidence

Answer 29

The number of new cases of a disease or new health-related events in a defined population within a specified period of time, the specified period of time being the same for each person.

Question 30

Cut-off point

Answer 30

Syn: cut point. An arbitrarily chosen value (though based as closely as possible on evidence of likely harm), such as the level of systolic and / or diastolic blood pressure, that is used to designate hypertension. The distinction between “normal” and “abnormal”.

Question 31

Death rate

Answer 31

Syn: MORTALITY RATE. The frequency with which death occurs among a designated population during a designated period, usually a calendar year. Calculated by dividing the number of deaths (the numerator) by the number of people at risk in the population in which the deaths are occurring.

Question 32

Degrees of freedom

Answer 32

A term used in STATISTICS to describe the number of independent comparisons that can be made between the variables in a study.

Question 33

Dependent variable

Answer 33

In statistics, a variable the value of which is affected by other variable(s).

Question 34

Experimental epidemiological studies

Answer 34

Studies in which exposure is determined as part of the study design. This includes randomised and non-randomised clinical trials. Syn: INTERVENTIONAL EPIDEMIOLOGICAL STUDIES

Question 35

Fatality rate

Answer 35

The death rate observed in a designated series of people affected by a simultaneous event, e.g., victims of a disaster, or by a specified condition.

Question 36

Haphazard sample

Answer 36

A group that is selected without regard for any rigorous method of sampling, without using either a random allocation procedure or a systematic non-random sampling method, such as every tenth name in a telephone directory.

Question 37

Hawthorne effect

Answer 37

Any effect on the behavior of individuals and groups due to knowing that they are under observation. Frequently this knowledge promotes behavior that those observed believe the observer(s) expect(s) or hope(s) to see, i.e., the effect is usually beneficial.

Question 38

Heterogeneity

Answer 38

The degree of “unlikeness” with a group or community (e.g. in a group of randomised controlled trials being analysed in a META-ANALYSIS).

Question 39

Incidence

Answer 39

The number of new cases of a disease or new health-related events in a defined population within a specified period of time.

Question 40

Incidence density

Answer 40

The number of new cases of a disease or new health-related events in a defined population within a specified period of time when the periods of observation for each individual are not the same. Incidence density uses person-time (e.g. person years or person months of exposure.

Question 41

Independent variable

Answer 41

In a research study, the characteristic, quality, or entity that is hypothesized to influence DEPENDENT VARIABLES that are being measured.

Question 42

Lifetime risk

Answer 42

The probability, often expressed as a percentage, that at some stage in life a person will develop a designated condition, specifically cancer. For example, there is a lifetime risk of about 25% of getting cancer, and a lifetime risk of about 20% of dying of cancer in upper-income countries.

Question 43

Loss to follow up bias

Answer 43

``` Erroneous conclusions (e.g. to a COHORT STUDY) as a consequence of LOSS TO FOLLOW-UP. The general term for all original participants in a COHORT or other long-term study who cannot or do not take part in further phase(s) for any reason, e.g., death, refusal, being untraceable, or having moved away from the study region. ```

Question 44

Masking

Answer 44

Syn: BLINDING. In clinical trials, the process of withholding the identities of study subjects in treatment and control groups, usually described as “blinding”. The word “masking” is preferred to “blinding” by some clinical trial specialists on the grounds that “blinding” may be an emotive word.

Question 45

Matching

Answer 45

The process of making a study group and a comparison group comparable with respect to extraneous (CONFOUNDING) factors (such as age, sex, ethnicity etc.).

Question 46

Measures of central tendency

Answer 46

A general term for several parameters related to the distribution of variables around some “central” value. These parameters are the MEAN, MEDIAN, and MODE. The median is the value at the point where exactly half the values in a set are higher and half are lower when all the values in a set are arranged in sequence. The mode is the most commonly occurring value in the set. The mean can be one of three kinds: the arithmetical mean is the sum of all the values in a set divided by the number of values in the set; the geometric mean is calculated by adding the logarithms of all the values in the set, dividing by the number of values, and taking the antilogarithm of this number; and the harmonic mean is the sum of the reciprocals of the values in a set, divided into the number of values in the set.

Question 47

Mortality rate

Answer 47

Syn: DEATH RATE. A general term for rates compiled from data on the number of deaths in relation to a specified population at risk. Several common variations are crude mortality rate, age-, sex-, and cause-specific mortality rates, and adjusted mortality rates. See also RATE, RISK and ADJUSTMENT.

Question 48

Multiple regression

Answer 48

A set of statistical techniques used, e.g., in sociological and epidemiological analyses, when many variables are involved. These tests commonly used include linear regression analysis and logistic regression analysis.

Question 49

Nominal data

Answer 49

Data that are identified in terms of their names (or numbers, used as labels rather than as numerical value) without regard for other qualities, such as rank and amount.

Question 50

Nominal scale

Answer 50

A classification method that organizes data according to specified qualitative criteria, such as country of birth, ethnic origin, without regard for any quantitative criteria, such as socioeconomic status.

Question 51

Nonparametric method

Answer 51

Syn: distribution-free method. A statistical method that can be used regardless of the manner in which the data to be analyzed are distributed.

Question 52

Normal distribution

Answer 52

A distribution that is symmetrical about a midpoint, with the largest number of values clustered closest to this and diminishing further away. Also called GAUSSIAN DISTRIBUTION.

Question 53

Null hypothesis

Answer 53

The proposal that one variable has no association with another or that two populations do not differ from one another.

Question 54

null hypothesis prediction

Answer 54

The predicted value of the odds ratio, relative risk or other effect size if the null hypothesis is true – one (unity) for odds ratio or relative risk and zero for the degree of difference, for example, in the quantitative effect on blood pressure of two drug interventions.

Question 55

Odds

Answer 55

The statistical calculation of the number of events as a fraction of nonevents or as a fraction of all that are theoretically possible. For instance, in the roll of a six-sided die, the odds that a specific number will roll to the top are 1 in 6.

Question 56

Odds ratio

Answer 56

Syn: cross-products ratio. Strictly speaking, the ratio of two odds, but as used in the analysis of data from a case control study, a simple calculation that yields an approximate value for the RELATIVE RISK of the exposure that has been examined in a CASE CONTROL STUDY. The odds ratio is best explained in a simple table The odds ratio is ad/bc. The odds ratio yields an approximation of RELATIVE RISK that is valid for rare and uncommon diseases but not for diseases with an incidence rate greater than about 1/100.

Question 57

Ordinal scale

Answer 57

A system of classification into ordered categories, e.g., a socioeconomic grouping according to income and/or educational level. Contrast NOMINAL SCALE.

Question 58

p, P

Answer 58

The abbreviation for the estimated probability of a result equal to or more extreme than that observed in a study, often appearing in publications together with the symbol < for “less than”. . Conventionally, a probability of less than 1 in 20, i.e., p<0.05, is considered unlikely to have occurred by chance in many situations arising in medicine, e.g., results of clinical trials, CONFIDENCE INTERVALS usually convey more information about the uncertainty attaching to an estimate than do p values, and are therefore a preferable way to express results of a study.

Question 59

Period prevalence

Answer 59

The number (or proportion) of individuals known to have a disease or health-related state in a given population during a specified time period.

Question 60

Point prevalence

Answer 60

The number (or proportion) of individuals known to have a disease or health-related state in a given population at a specified point in time.

Question 61

Poisson distribution

Answer 61

A distribution function to describe rare events such as cases of uncommon diseases.

Question 62

Precision

Answer 62

The quality of being exactly or sharply defined. For example, a measurement of length expressed to decimal places of millimetres is more precisely defined than one given to the nearest metre. To be distinguished from ACCURACY.

Question 63

Prevalence

Answer 63

The number (or proportion) of individuals known to have a disease or health-related state in a given population.

Question 64

Proportional morality ratio (PMR)

Answer 64

The number of observed deaths from a specified cause in a defined population divided by the number of deaths that would be expected in a standard population, with both numbers expressed as a proportion of all deaths. The PMRs of specific occupational groups are sometimes used to determine whether their mortality experience differs significantly from that of the general population, justifying further investigation.

Question 65

Quota sampling

Answer 65

A sampling method in which the proportions vary in designated subgroups according to such criteria as age group, occupation, social class etc.

Question 66

Random allocation

Answer 66

Allocation of persons or entities, such as therapeutic regimens, by a formal process using an accepted method of random sampling, such as a table of random numbers. Used in RANDOMISED CONTROLLED TRIALS.

Question 67

Random error

Answer 67

The portion of the variations of measurements and other observations that is due to the operation of chance alone. Random error increases in size and importance if sampling methods are poorly chosen and response rates are low.

Question 68

Random sample

Answer 68

Syn: probability sample. A sample drawn by a process of RANDOM ALLOCATION from a population in a manner aimed at ensuring representativeness.

Question 69

Rate

Answer 69

A measure of the frequency with which specified events occur in a defined population in a designated period of time. Rates have three components: a numerator, the number of events; a denominator, the numbers in the population at risk of experiencing the event; and a time dimension, the period over which the events occur.

Question 70

Ratio

Answer 70

The result of dividing one quantity by another without regard for details such as time dimension. Rates, proportions, and percentages are all types of ratios. The distinction between a proportion and a ratio is that, whereas the numerator of a proportion is included in the population defined by the denominator, this is not necessarily so for a ratio, which expresses the relationship of two separate and distinct quantities, neither of which is included in the other. See also PROPORTION and RATE.

Question 71

Relative risk

Answer 71

Syn: risk ratio, hazard ratio. The ratio of the risk of disease or death among the exposed group to the risk among the unexposed. I.e. the incidence of the disease or death in the exposed group divided by the incidence in the unexposed.

Question 72

Response (rate)

Answer 72

The proportion, percentage, or fraction of individuals who reply to inquiries that have been directed to all the individuals in a population, or a representative sample of them. The VALIDITY of findings or results and conclusions derived from them is dependent on the response rate. It the response rate is low, i.e., if there are large proportions of nonresponders, the VALIDITY of the results is dubious, even if impeccable sampling methods have been used, because of the magnitude of random error.

Question 73

Standard error

Answer 73

The standard if deviation of an estimate. The term often refers to the standard deviation of the observed mean, represented by the symbol σ and calculated from a sample and the true population mean.

Question 74

Standard deviation

Answer 74

The most widely used measure of spread or dispersion from the central value of a frequency distribution. It is the positive square root of the variance from the mean or central value of the individual measurements in the distribution.

Question 75

Standard population

Answer 75

A population used for STANDARDIZATION. Any population can be used for this purpose provided its age and sex structure are precisely known.

Question 76

Standardisation

Answer 76

A set of techniques (DIRECT STANDARDISATION and INDIRECT STANDARDISATION) used to remove as much as possible the effects of differences in confounders (such as age, sex, ethnicity, social class) when comparing two or more populations.

Question 77

Standardised mortality ratio (SMR)

Answer 77

The ratio of the number of deaths observed in the study group or population to the number that would be expected if the study population had the same stratum specific rates as the standard population. Usually expressed as a percentage.

Question 78

Standardized incidence ratio

Answer 78

The ratio of the incident number of cases of a condition in the study population to the number that would be expected if the study population had the same incidence rate as a standard population. This is an infrequently used parameter.

Question 79

Stratification

Answer 79

The process of or result of separating a sample into several sub samples according to specified criteria (usually of a confounder such as age, sex, ethnicity etc.).

Question 80

Student’s t test

Answer 80

A statistical test for the degree of similarity or difference between two or more distributions, invented by the British mathematician W.S. Gossett, who published under the pseudonym “Student” because of the conditions of his employment. Cf. Chi square

Question 81

t-distribution

Answer 81

A mathematical model that defines the relationship between two or more sets of data, used to determine whether they are significantly different.

Question 82

Weighting

Answer 82

A method of adjustment to allow for differences between the study population and a “standard” one, between two or more populations that are being compared, or among any set of variables in which confounding factors might otherwise invalidate comparisons. It can be done in several ways, often by taking different predetermined proportions from several parts of the study population or set of variables, to ensure that numbers in each are sufficient to yield stable rates.