The Test of Financial Literacy: Development and measurement characteristics

ABSTRACT

The Test of Financial Literacy (TFL) was created to measure the financial knowledge of high school students. Its content is based on the standards and benchmarks stated in the National Standards for Financial Literacy (Council for Economic Education 2013). The test development process involved extensive item writing and review. Test data collected from 1,218 high schools to evaluate the measure indicate that the overall test is reliable and valid, and test items contribute to the effectiveness of the instrument. Further test analysis was conducted using an item response theory (IRT) model with four parameters to estimate item discrimination, item difficulty, guessing, and inattention. The IRT results indicate that the measure is effective in assessing student financial literacy across a broad range of student abilities.

KEYWORDS:

• Financial literacy
• four-parameter IRT model
• item response theory
• personal finance
• testing
• validity

JEL CODES:

• A21
• D14

Notes

1. All three tests (TFL, TFK, and BFT) and the examiner's manuals are available online from the CEE at http://councilforeconed.org/resources/online-assessment-center/.

2. Members of the national writing committee were William Bosshardt, Professor of Economics and Director of the Center for Economic Education at Florida Atlantic University; Elizabeth Breitbach, who at the time of this work was a Clinical Assistant Professor of Economics at the University of South Carolina; Brenda Cude, Professor in the Department of Financial Planning, Housing, and Consumer Economics at the University of Georgia; Andrew Hill, an economic education advisor at the Philadelphia Federal Reserve Bank; and Bonnie Meszaros, Associate Director of the Center for Economic Education and Entrepreneurship at the University of Delaware. William Walstad was the project director and Ken Rebeck the associate director. The CEE representative for the project was Kevin Gotchet.

3. Further investigation of that item, however, did not reveal any particular statistical or content problem. A graph of the item characteristic curve showed the typical S-shape and had no unusual features. From an observed score perspective, the item was challenging, but the item difficulty properly sorted students with and without instruction (0.41 versus 0.37). The corrected item-total correlation (0.35) was reasonably high. The financial literacy content appears to be sound.

4. IRT estimates for item difficulty (table 4) also are correlated highly (−0.911) with estimates of item difficulty based on observed scores (table 3).