States may use a variety of target setting approaches^{1} and should clearly and completely explain their rationale and methods. The following information provides an overview and examples of methods states may want to consider.
Percent or Percentage Point Improvement
Percent or percentage point improvements are common methods for setting targets. Following are several different ways of determining and applying these changes over time to targetsetting methods.
Read More +
Average yearoveryear growth/change. Using historical data, calculate the average growth/change from year to year. This change can be calculated as a percent change/improvement or a percentage point improvement. For example, if the average growth is 0.8 percentage points, add that amount to the current data for each year.
Table: Percentage Points (using child outcomes data)

Actual 
Targets 

2015 
2016 
2017 
2018 
2019 
2020 
2021 
2022 
2023 
2024 
2025 
SS1 socemo 
70.0 
70.8 
71.0 
71.8 
72.8 
74.0 
74.8 
75.6 
76.4 
77.2 
78.0 
Difference 

0.8 
0.2 
0.8 
1.0 
1.2 

Average Actual Growth = 0.8 
The example below uses the average percent change from year to year and applies that to each future year. Percent change is less commonly used than percentage point change and is more difficult to explain.
Table: Percent Change (using child outcomes data)

Actual 
Targets 

2015 
2016 
2017 
2018 
2019 
2020 
2021 
2022 
2023 
2024 
2025 
SS1 socemo 
70.0 
70.8 
71.0 
71.8 
72.8 
74.0 
74.8 
75.7 
76.5 
77.4 
78.2 
Percent Change 
1.14 
0.28 
1.13 
1.39 
1.65 


Average Percent Change = 1.12 
Overall growth/change. Calculate the overall growth from two historical points in time, e.g., from year 1 to year 5, using either percent or percentage point improvement. Increase the end target for five years out by that total growth. For example, if the total growth from 2015 to 2020 is 4 percentage points, add that to the end target five years out. Decide if the targets for each intervening year should increase incrementally by .80 percentage points (4 divided by 5 years) or in other increments, depending on state circumstances, e.g., the status of improvement initiatives.
Table: Percentage Points (using child outcomes data)

Actual 
Targets 

2015 
2016 
2017 
2018 
2019 
2020 
2021 
2022 
2023 
2024 
2025 
SS1 socemo 
70.0 
70.8 
71.0 
71.8 
72.8 
74.0 




78.0 
Overall difference from 2015 to 2020 = 4.0 
Moving (rolling) average. If the historical data are not stable, a moving average can be calculated and added to each of the future years. The moving average may be based on a period of two, three, or four years, depending on the number of years of historical data available. Rolling averages need to total numerators and denominators separately first and then calculate percentages
New baseline (or no historical data). If historical data are not available or if a new baseline has been established, e.g., due to changes in data collection methods, start with the new baseline (or most recent year of actual data) and increase that by a certain percentage or number of percentage points each year. For a percent improvement, the baseline is multiplied by a specific percentage, and the resulting value is added to or subtracted from the baseline. For a percentage point improvement, the baseline, itself a percentage, is improved by adding or subtracting a specific value, also known as a percentage point.
Start with the End Goal
Decide on the target for the last year of the SPP/APR cycle. One approach to setting the end goal is to determine a meaningful/statistically different value from baseline or current data. The Child Outcomes YeartoYear Meaningful Differences Calculator for States can be used for the C3/B7 indicators on child outcomes.
Read More +
The meaningful difference calculator uses an accepted formula to determine whether the difference between two percentages is statistically significant (or meaningful). Statistical tests of significance can be used to determine meaningful differences for other indicators. Once the end goal is set, determine the incremental targets for the intervening years.
Trend Analysis and Forecasting
A trendline, also referred to as a line of best fit, is a straight or curved line on a chart that shows the general pattern or overall direction of the data. Trend analysis is most often used to show data movement over time, particularly to estimate data in future years. You can decide on a target based on the trendline projection. An important consideration in trend analysis is how far back to go; that is, when to start the trendline.
Read More +
Tools, like Excel, can be used to add a trendline to a chart and extend the trendline to future years (forecast). Excel provides different options for trend analysis and forecasts, depending on the type of data you have.
 Linear. A linear trendline is used with simple linear data sets; that is, the pattern in the data points resembles a line. A linear trendline usually shows that something is increasing or decreasing at a steady rate.
 Moving average. A moving average trendline smooths out fluctuations in the data to show a pattern or trend more clearly. A moving average trendline uses a specific number of data points, averages them, and uses the average value as a point in the trendline. You can determine the number of data points to use in the moving average, e.g., two, three, or four.
 Logarithmic. A logarithmic trendline is a bestfit curved line that is most useful when the rate of change in the data increases or decreases quickly and then levels out.
Linear data typically require fewer data points for projections than other options. Excel can also create confidence intervals and display the Rsquared value of a trendline, which is a number that indicates how well your trendline fits your data. The closer the Rsquared value is to 1, the better the fit of the trendline.
Go to Trend Analysis or Forecasts for more information on using these Excel functions.
Statistical Modeling/Analysis
Statistical analysis can be used to help predict future results and thus, targets, using additional data such as population data, regional data, or outliers.
Read More +
For example, a state could stratify its data by the size of the local program/district and weight those data accordingly, or it could look at the change in the results of local programs/districts that have implemented improvement initiatives versus those that have not, and set targets based on scaleup plans.
Additional Considerations
For each of the approaches, consider changes in state circumstances that may impact performance in any given year, such as data quality issues or the scope and status of improvement initiatives. There may be legitimate reasons for maintaining stability for a few years, and targets may remain the same for several years.
Read More +
Similarly, targets in the intervening years may increase incrementally, but not by the same amount each year. However, targets must show improvement from the baseline in the end.
States may want to consider and use more than one method and bring the results of those methods to stakeholders for review and discussion. An effective way to engage stakeholders in the targetsetting process is to present these multiple options for targets, explain the rationale for each, and solicit feedback. Presenting these options visually, e.g., all on one graph, allows stakeholders to see the impact of each. An example is presented in Figure 8 in the Examples of Data Visualization section.
^{1}Hubbard, K., Makram, T., Klein, R., & Huang, D. 2020. TargetSetting Methods in Healthy People 2030. Healthy People Statistical Notes.