Chapter 3, Problem 1E

In addition to the key words, you should also be able to define each of the following terms:

Positive relationship	ratio scale
Negative relationship	self-report measure
Accuracy	physiological measure
Scale of measurement	behavioral measure
Nominal scale	range effect
Ordinal scale	artifact
Interval scale	subject roles or subject role behaviors

To determine

The definition of the following terms.

Explanation of Solution

Positive Relationship:

In statistics, a positive relationship or positive correlation is defined as the relationship between two variables in which, increment (decrement) in one variable results in increment(decrement) of another variable (not necessarily in same ratio). In other words, positive relationship is the relationship between the two variables in which both variable moves in same direction.

For example, increment in RAM results in increment in price of mobile. The positive relationship can also be seen graphically through scatter plot.

Negative Relationship:

In statistical terms, a negative relationship or negative correlation is defined as the relationship between two variables in which both the variables move in opposite direction, that is, increment(decrement) in the value of one variable results in decrement (increment) of other.

For example, on increasing the temperature of microwave, the time taken by water to boil is decreased.

Accuracy:

Statistics is not about deterministic value, it's about estimated value which is close to accurate value. While doing an experiment, one can never get exact value. So accuracy tells the closeness between true value and estimated value.

For example, a sample is drawn from a population having mean 30 and the sample mean is 29 then one can say that sample mean is accurate to within 1 unit.

Scale of Measurement:

The term 'measurement' is used in our daily lives very broadly but in statistics, the term 'scale of measurement' is used which is more appropriate and is used to categorize the variables. The scale of measurement gives some ways to categorize the variables. The scale of measurement is divided into four scales- nominal, ordinal, interval and ratio scale. Each scale has its own properties on the basis of which the variables can be categorized. This categorization is very useful while performing statistical analysis and without this categorization, it is not possible to perform analysis. The scale of measurement works as base of statistical analysis.

Nominal Scale:

The scale which consists of variables having categories and do not have any order is known as nominal scales and the variables are known as nominal or categorical variables. These variables are simply used as identifiers such as blood group (AB, A, B, O), gender (male and female), residential area (urban and rural) etc.

In this, categories of any variables can be assigned the codes in the form of the numbers and those numbers work as the labels. For a variable gender, male can be assigned as 0 and females as 1.

Ordinal Scale:

The ordinal scale contains the variables having categories and should be placed in order and the variables are known as ordinal variables. For example, ranks achieved in drawing competition, that is, first, second and third positions.

This scale is not used to represent the quantity.

Interval Scale:

The term 'interval' means 'gap in between' which shows the importance about each value between items. An interval scale is a scale in which zero point is arbitrary. In this scale, scores can be added and subtracted but cannot be multiplied meaningfully. For example, temperature because there is no such thing as 'no temperature'.

Ratio Scale:

The scale, consisting of the variable having true zero point is known as the ratio scale. In ratio scale, the percentage is useful unit. For example, weight is expressed in kilograms and also in grams and they are linearly related as 1kg=1000gms. Here 0 kilograms and 0 gram is same and would have no weight.

Self-Report Measure:

In self-report measure, we do a study based on questionnaire or survey. Here, respondents select their responses themselves and questions are based on their feelings, attitudes, behaviour etc. and record their answers.

Self report measures are operational definitions in which respondents are asked to fill the questionnaire about their own behaviour or mental conditions.

Psychological Measurement:

Psychological measurement deals with the measurement of psychological traits or mental abilities like intelligence, aptitude, interest, opinion, honesty or attitude. Psychological characteristics can be measured only with some degree of unreliability. The psychological scale is an interval scale because there is no absolute zero-point on the scale.

Behavioural Measure:

Behavioural measures are tests and methods, targets on information from a subject of interest. It contain over actions and reactions of respondents.

Range Effect:

When each subject of interest receives more than one experimental condition then range effect occurs. The range effects are critical to measurement of almost any kind of human behaviour.

Artefact:

In econometrics, artefact is defined as an error which occurs during data manipulation or in the case of selection of faulty choice of variables.

Subject role or subject role behaviours:

It includes the study of human thoughts and actions, individually as well as collectively. In this, the behaviour and reactions of human are recorded for any particular situation.