Explore and Analyze Data41% of the exam
Trend lines and forecasting models
Add and edit trend lines (model types, factors, per-color lines, stacked-bar limits) and judge them with R-squared and p-values, then add a forecast with default settings, clear the view's forecasting constraints and customize length, source data, model and prediction intervals in Forecast Options.
Summarize, model, and customize data by using the Analytics pane: Totals and subtotals; Reference lines; Reference bands; Average lines; Trend lines; Distribution bands; Forecast by using default settings; Customize a data forecasting model.
~8 min read · 10 practice questions ·
What the exam expects
- Add a trend line from the Analytics pane, pick its model type, and fix a view where trend lines do not appear.
- Edit trend lines: remove a field as a factor, turn off one line per color, force the y-intercept to zero, show confidence bands.
- Judge a trend line with R-squared and p-values from the tooltip and Describe Trend Model.
- Add a forecast with default settings, and remove whatever blocks it (table calculations, totals, disaggregated measures, exact dates).
- Customize a forecast in Forecast Options: length, Aggregate by, Ignore last, model (Automatic or Custom additive/multiplicative) and prediction interval.
- Totals, reference lines, bands, average lines and distribution bands: see Totals, reference lines, bands and distributions.
Key terms:
- Factor: a dimension in the view that splits the trend line model into separate lines.
- p-value: the probability that the trend equation came from random chance; smaller is more significant.
- Ignore last: trailing periods excluded from estimating the forecast model.
- Prediction interval: the shaded band the forecast value is expected to fall in at a set percentage.
Key facts
Trend line axes
- Detail
- Both must hold a field interpretable as a number
Stacked bars
- Detail
- No trend lines; clear Analysis > Stack Marks
Confidence bands
- Detail
- 95% by default; not for Exponential
Linear R-squared
- Detail
- Equals the square of CORR
Significant p-value
- Detail
- 0.05 or less is often considered sufficient
Forecast blockers
- Detail
- Table calcs, disaggregated measures, percent calcs, totals, Exact Date
Web editing
- Detail
- Forecast visible, but cannot be added or modified
Minimum data
- Detail
- 5 points for trend; 24 for a 12-month season
Finer data
- Detail
- Only for SUM or COUNT measures
Ignore last default
- Detail
- Trims one period
Multiplicative limit
- Detail
- No values less than or equal to zero; no multiplicative trend with additive season
Band width
- Detail
- Lower percentile gives narrower prediction bands
Sources for this section (5)
- Add Trend Lines to a Visualizationhelp.tableau.com (opens in a new tab)
- Forecastinghelp.tableau.com (opens in a new tab)
- How Forecasting Works in Tableauhelp.tableau.com (opens in a new tab)
- Configure Forecast Optionshelp.tableau.com (opens in a new tab)
- Create a Forecasthelp.tableau.com (opens in a new tab)
- Why does Tableau not offer confidence bands for an Exponential trend line?
- When is a date part valid for forecasting?
- What does the shaded area on a default forecast show?
- Why does Tableau refuse a multiplicative trend with an additive season?
The questions are free to read. Answers with their sources and self-grading are included with the pack.
Keep reading with the Tableau Data Analyst pack
- Core concepts
- Exam traps
- Worked scenario
The pack includes:
- All 30 lessons of the Tableau Data Analyst guide, with reading progress
- Interactive decision walks and step-by-step sequence steppers
- Practice exams purchase adds inline bank questions and topic practice (10 questions for this lesson)
- An official source cited for every claim
- An AI tutor that answers from the guide
- Choose guide-only, practice-only, or save $4 with both