Chapter 2 - Forecasting data Flashcards
TIME SERIES
Is a series of figures or values recorded over time.
TREND
Is the general movement of a time series over a long period of time. It can often be attributed to the impacts of sales growth/decline or inflation.
If the trend was plotted on a graph it would show as a smooth line or curve.
SEASONAL VARIATION
Is a predicted movement away from the trend. These are due to repetitive events which occur over a short but fixed period of time.
E.g. more icecream sales in the summer than winter.
CYCLICAL VARIATION
Is a recurring pattern over a longer period of time, but not generally of a fixed nature - unable to be predicted with certainty.
E.g. movements away from the trend due to variations in the economy - recession to economic growth
RANDOM VARIATION
Is an irregular variation due to the rate or chance occurrences. They are beyond control of the organisation.
E.g. COVID, hurricane, flood affecting the factory/warehouse.
E.g. high sales of a newpaper due to the exclusive celebrity pictures.
INDEX NUMBERS
An index measures over time the average changes in the values, prices or quantities of a group of items.
E.g. Retail price index, which measures the changes in the costs of items of expenditure of the average household.
RPI Characteristics:
BASE PERIOD
A point in time with which current prices/quantities are compared. The index number in the base year
RPI Characteristics:
BASKETS
These are the items that are included within the index, of which the movement in their price is analysed within the index.
E.g. the items measured in the RPI include food, clothing, head & light, and restaurants and hotels.
RPI Characteristics:
WEIGHTINGS
Are used to give the relative importance of each item. In RPI, each item is given a weighting depending on the proportion it constitutes towards total household expenditure.
RPI Characteristics:
Index numbers
Show the general movement of the data over a period of time and may be used to inflate costs to future periods for forecasting, or deflate costs to a previous period for comparison.
INTERPOLATION
Is forecasting data within the historical data range.
EXTRAPOLATION
Is forecasting data outside of the historical data range.
Product life cycle:
Development
Revenue: No external market, therefore no revenue generated
Cost: High level of capital development costs. E.g. Purchase of non-current assets
Product life cycle:
Introduction
Revenue: Initial demand is low, so revenue generated is also low.
Cost: Advertising, promotion & continued development costs. E.g. Direct costs
Product life cycle:
Growth
Revenue: Demand for the product leading to increased sales revenue
New competitors enter the market attracted by the success and profit of the product, putting pressure on our sales revenue.
Cost: Fixed promotion costs remain high. Direct costs and overheads become the biggest proportion of cost as production is increased to satisfy demand.
Product life cycle:
Maturity
Revenue: Demand peak becomes steady. Revenue generation is at it’s highest, as the product is established and have customer loyalty.
Cost: Mechanisation of the production process. Bulk buying discounts. Skilled staff meaning more efficiencies. Lower fixed production overheads spread over more units.
Product life cycle:
Decline
Revenue: Market saturation reached and eventual drop in demand.
Cost: Production cost reduce. Increased obsolescence cost (e.g. lost contribution on discounted products, wastage costs of unsold inventory)
Random sampling
Can be used when the entire population being considered is known. Random numbers are used to select a sample from the population.
Stratified sampling
Is a type of random sampling which seeks to divide (or stratify) a population into groups , and then selects a random sample from each group based on the proportionate size of each group.
This is done by dividing the number in each group by the total population, and then multiplying this fraction by the overall sample.
Quota sampling
Is used in situations where a number of different groups of the population can be identified. The number of samples required from each group is then determined and the data is taken from that required number in a non-random manner.
