Econometrics: Literally, the word econometrics means economic measurement or the measurement of economic relationships. According to Goldberger: Econometrics may be defined as the social science in which the tools of economic theory, mathematics and statistical inference are applied to the analysis of economic phenomena.
Uses/functions/objectives of econometrics:
Econometrics is often used to
- Testing economic theories and models
- Estimating numerical values of economic coefficients
- Forecasting future events
- Evaluating programs and policies
- Identifying cause and effect relationships
Mathematical economics vs econometrics: Mathematical economics concerns expressing economic theory in mathematical form (equations) without measurability or empirical verification of the theory. It expresses the economic relationship in exact or deterministic form. Whereas econometrics empirically tests economic theory or hypothesis and provides estimated values to economic relationships.
Economic statistics vs econometrics: Economic statistics is mainly concerned with collecting, processing, and presenting economic data in the form of charts and tables. Whereas econometrics uses this data to empirically test economic theory or hypothesis. Moreover, econometrics is concerned with causal inference while statistics is mostly concerned with statistical inference.
Cross-sectional data consists of observations collected from various entities such as individuals, households, firms, cities, states and countries at a single point in time. For example, GDP of Asian countries for the year 2023, number of deaths due to coronavirus pandemic in the year 2020. It is often denoted by subscript i.
Time series data consists of observations collected over multiple time periods for a single entity. For example, data about Real GDP, Inflation, Unemployment and Life expectancy of Pakistan from 1991 to 2019. It is often denoted by subscript t.
Pooled (or combined data) have features of both cross section and time series data in which each cross-section unit may not be the same for each time period. For example, data of different students over different semesters and each student may not be the same in each semester.
Panel (or longitudinal) data is a combination of cross-section and time-series data in which data on the same cross-sectional units are collected over multiple time periods. For example, data about GDP, inflation, unemployment rate, money supply, and investment for all developing countries from 1970 to 2023. It is often denoted by subscript it.
Balanced vs unbalanced panel data: In balanced panel number of time observations are same for all cross-section units. There are no missing observations. In unbalanced panel number of time observations are not same for all cross-sectional units. There are missing observations for each cross section.
Primary data is the data collected for the first time by a researcher for his/her specific research purpose. It is collected through surveys, interviews, experiments, observations, and questionnaire. For example, a student collects data from college students to study the relationship between laptop distribution and exam scores.
Secondary data is data that has already been collected by an institution or researcher for different purposes. it can be obtained from sources such as books, reports, articles, online databases and surveys. For example, a student collects data from WDI to study relationship between inflation and unemployment.
Experimental data is collected through controlled experiments where researchers can manipulate one or more independent variables to observe cause-and-effect relationships. For example, testing the effectiveness of a new vaccine.
Non-Experimental data or observational data is collected by observing and recording events, behaviors, or phenomena as they naturally occur without manipulation. This method is used where experiments are not possible, not ethical or expensive. For example, studying the relationship between consumption and income.
Quantitative data is data that can be measured numerically. Such as GDP, GNP, Exports, Prices, Investment etc. Quantitative variables can be classified into two broad categories namely: (i) discrete variables and (ii) continuous variables
Qualitative data or categorical data is the data which cannot be measured numerically but can be classified into several groups or categories. . Such as gender, education level, social status, religion, race etc Categorical data can be nominal or ordinal.
Nominal scale is the lowest level of measurement scale where data is classified into mutually exclusive qualitative categories without ordering or ranking. Examples are gender, colors, house numbers, blood group etc. It is often represented by bar charts or pie charts.
Ordinal scale has the characteristics of nominal scale and in addition has the property of ordering or ranking of categories but the difference between them is not meaningful. For example, the performance of students in class test like excellent, good, fair, poor, very poor or classification of countries based on GNI per capita.
Interval scale has all the characteristics of ordinal level data and the difference between values is also meaningful but the ratios not. It does not have true zero point. For example, temperature recorded for a city over different months.
Ratio scale is the highest level of measurement scale which has all characteristics of interval scale and the ratio between values is also meaningful. It has also a true zero point. For example, income, prices, weight, height, money, output etc.
