A time series is a set of sequentially recorded numerical data points for a given variable, arranged in sequential order to track variation.
A. The sum of the weights in a moving average is zero.
B. If the growth rate is constant, the trend is a polynomial (quadratic).
C. Time series data analysis is of three types: time series, cross-sectional and pooled data.
D. When a time series is a collection of data for multiple variables and how they vary over time, then it is known as a multivariate data set.
E. The patterns and behaviour of the data in any time series are based on three components.
Choose the correct answer from the options given below:
Answer & explanation
Correct answer: option 4
The correct answer is Option (4) → C, D only
The facts needed about time series:
\(\text{k-period moving average: each value has weight } \frac{1}{k},\ \text{total } k \times \frac{1}{k} = 1\)
\(\text{constant growth rate: } y = ab^t,\ \text{an exponential trend}\)
Data analysed over time are of three types: time series data (one variable at successive times), cross-sectional data (one or more variables at the same time) and pooled data (a combination of the two). A series of several variables varying over time is a multivariate series. A time series has four components: secular trend, seasonal, cyclical and irregular.
So the weights of a moving average add up to 1, a constant growth rate gives an exponential trend, and a time series has four components.
Checking the statements
(A) The sum of the weights in a moving average is zero. — Incorrect. The weights add up to 1 (each is \(\frac{1}{k}\)).
(B) If the growth rate is constant, the trend is a polynomial (quadratic). — Incorrect. A constant growth rate gives an exponential trend.
(C) Time series data analysis is of three types: time series, cross-sectional and pooled data. — Correct. These are the three types of data analysed.
(D) When a time series is a collection of data for multiple variables and how they vary over time, then it is known as a multivariate data set. — Correct. This is the definition of a multivariate series.
(E) The patterns and behaviour of the data in any time series are based on three components. — Incorrect. There are four components: trend, seasonal, cyclical, irregular.