Making a prediction from a trend

Name:

Date:

Class:

Quick Questions

Shade in the A or B circle for each question.

1. A chart shows a school canteen's weekly sales rising steadily. You extend the trendline to guess next week's sales. What is this?

A A guarantee

B A prediction

2. Which statement about a prediction from a trendline is true?

A It is an estimate that could be wrong

B It is a fact that is always right

3. A trendline on a chart of Year 7 reading minutes goes up each week. Which is the best way to make a prediction for next week?

A Extend the trendline forward one step

B Pick any number you like

4. Your chart shows 6 months of data about a phone's battery health. Which prediction is likely to be MORE reliable?

A Predicting 5 years from now

B Predicting next month

5. Why is it risky to use 4 weeks of plant growth data to predict the plant's height in 2 years?

A The trend may change over a long time

B Trendlines stop working after 4 weeks

Match the Pairs

Draw a line connecting each item on the left to its match on the right.

Left column		Right column
1Forecast forward 3 periods		APredictions get less reliable the further ahead you go, so 100 periods is far less accurate than 1.
2=FORECAST(13,B2:B13,A2:A13)		BPredicted sales next month: about 120
3A prediction is an estimate based on a trend, not a certainty, so we say 'about' or 'around'		CA prediction is treated as a certainty. It should say the sales are predicted to be about 120.
4Predicting far past the known data is less reliable than predicting just one step ahead		DForecast 1 period = more reliable, Forecast 50 periods = risky
5Next month's sales will be exactly 120.		EExtends the trendline 3 time periods past the last data point to estimate future values
6Trendline forecast: 100 periods ahead is just as accurate as 1 period ahead		FPredicts the value for period 13 using the pattern in the existing data

Definitions

prediction

Fill in the Blanks

Complete each line of code by filling in the missing parts.

=________(6,B2:B6,A2:A6)

Weeks 1 to 5 are in A2:A6 =FORECAST(_,B2:B6,A2:A6)

Trendline: y = 2x + 10 =_*6+10

Trendline: y = 2x + 10 =2*6_10

A7 holds week 6 =FORECAST(__,B2:B6,A2:A6)

Spot the Error

Shade in the letter of the line that contains an error.

Question 1

A A2:A6 = month, B2:B6 = sales

B =FORECAST(6,A2:A6,B2:B6)

C Result is an estimate only

Question 2

A Scores: 12, 16, 20 (weekly)

B step ← 8

C last ← 20

D next ← last + step

Question 3

A A2:A7 = year, B2:B7 = height

B Known years are 1 to 6

C Predict height for year 7:

D =FORECAST(6,B2:B7,A2:A7)

Question 4

A Known data: days 1 to 10

B Day 11: close to the data

C Day 100: just as reliable

D Far predictions are riskier

Question 5

A Right-click the trendline

B Choose Format Gridlines

C Forecast Forward: 2 periods

D Line extends past the data

Answer Key

Making a prediction from a trend

Section A: Quick Questions

B — A prediction

A — It is an estimate that could be wrong

A — Extend the trendline forward one step

B — Predicting next month

A — The trend may change over a long time

Section B: Match the Pairs

Forecast forward 3 periods → Extends the trendline 3 time periods past the last data point to estimate future values

=FORECAST(13,B2:B13,A2:A13) → Predicts the value for period 13 using the pattern in the existing data

A prediction is an estimate based on a trend, not a certainty, so we say 'about' or 'around' → Predicted sales next month: about 120

Predicting far past the known data is less reliable than predicting just one step ahead → Forecast 1 period = more reliable, Forecast 50 periods = risky

Next month's sales will be exactly 120. → A prediction is treated as a certainty. It should say the sales are predicted to be about 120.

Trendline forecast: 100 periods ahead is just as accurate as 1 period ahead → Predictions get less reliable the further ahead you go, so 100 periods is far less accurate than 1.

Section C: Definitions

prediction — Extending a trend past the known data, such as running a trendline forward, to estimate a future value. It is an estimate, not a certainty, and gets less reliable the further ahead you go.

Section D: Fill in the Blanks

=FORECAST(6,B2:B6,A2:A6)

Weeks 1 to 5 are in A2:A6
=FORECAST(6,B2:B6,A2:A6)

Trendline: y = 2x + 10
=2*6+10

Trendline: y = 2x + 10
=2*6+10

A7 holds week 6
=FORECAST(A7,B2:B6,A2:A6)

Section E: Spot the Error

B — FORECAST needs the known y values (sales) before the known x values (months). The ranges are swapped. It should be =FORECAST(6,B2:B6,A2:A6).

B — The scores go up by 4 each week (12, 16, 20), so the step in the trend is 4, not 8. It should be step ← 4.

D — Year 6 is already in the known data, so this is not a prediction. To predict the next year, the first value should be 7.

C — The further ahead you predict, the less reliable the estimate is. Day 100 is far from the known data, so it should say much less reliable.

B — The Forecast option is in the Format Trendline menu, not Format Gridlines. Gridlines settings cannot extend a trend forward.