FASTfact: Spanish Benchmark and Norms
As defined in the FastBridge Learning Benchmarks and Norms Interpretation Guide (2016), standards of performance require the use of both data analysis and expert judgment. In general, FAST benchmarks predict proficiency on nationally normed assessments and state-level accountability tests. There is substantial variability in the definition of proficiency across states, so the default expectation of proficiency on most FAST measures approximates the 40th percentile. The high-risk benchmark approximates the 15th percentile. State-specific custom benchmarks can be developed at the request of users. Users may also develop and customize their own benchmarks.
The norms and benchmarks on Spanish measures have been more difficult to develop and refine, in part, because performance in Spanish does not predict performance on English accountability tests. As a result, developers used performance on a nationally normed measure of broad reading in Spanish along with the expert input of researchers and native Spanish-speaking teachers.
Materials and Benchmarks
The initial materials and benchmarks for assessment in Spanish were developed at the University of Minnesota in 2013 and 2014. The research and development team included Dr. Theodore Christ, Dr. Lori Helman, native Spanish-speaking teachers from English-to-Spanish immersion settings, and Spanish-speaking doctoral students. The materials are not translations of the English assessments. Instead, they were originally developed as Spanish measures with specifications that were substantially similar to the English materials whenever appropriate.
Validation and benchmark studies were established with the Aprenda-3 (Pearson, 2005). These studies were conducted to determine what level of reading automaticity in Spanish predicts success on broader Spanish measures that include vocabulary and reading comprehension. The Aprenda-3 is a culturally inclusive, group-administered, standardized test developed by Hispanic educators who modeled it on the Stanford Achievement Test-10 (Pearson, 2005). It test was normed on Spanish-speaking students from the United States, Mexico, and Puerto Rico in the spring and fall of 2004. Criterion validity ranged from about .40-to-.75 across measures. The Spanish CBMreading and earlyReading composites generally met the highest standards of the National Center for Response to Intervention for both diagnostic accuracy and criterion validity.
The initial results for earlyReading Spanish indicated a mid-year composite of 48, which corresponded with 15 Letter Sounds, 13 Onset Sounds, 10 Syllables, and 18 Word Segments. The initial results for CBMreading Spanish indicated mid-to-late year standards of 100, 106, 107, 132, and 134 for grades first to fifth respectively. Those standards predicted proficiency at the 40th percentile of international norms for native Spanish-speaking students on the Aprenda-3. As with CBMreading English, FAST CBMreading Spanish benchmarks are higher than those for other passage sets. This is because of the specifications used for passage development.
FAST oral reading passages were written with lower text complexity compared to other passage sets. This optimizes the validity of automaticity/fluency scores. More-skilled readers tend to slow down as text complexity increases while less-skilled readers, who do not monitor their comprehension, do not slow down. That confounds the measurement of passage reading automaticity/fluency because lesser-skilled readers over-perform and more-skilled readers under-perform. That is why (a) passages were constructed with less text complexity, (b) students read more words per minute on FAST passages, and (c) benchmarks tend to be higher for FAST CBMreading.
Automatic Passage Reading
Automatic passage reading is a foundational skill, which is necessary—but not sufficient—for comprehension to occur during reading. This widely used and well-established fact is founded in substantially convergent theories and evidence, which include the Simple View of Reading (Savage, 2001), Information Processing Model (Palmer et al., 1985), Automaticity Theory (LaBerge & Samuels, 1974), and Unitization Theory (Ehri, 2005). This applies to both English and Spanish reading.
Here is why. It is typically necessary to automate the lower-level skill of word identification before higher-order comprehension skills are established and consistent (LaBerge & Samuels, 1975). The reason for this is that humans have restricted cognitive capacity. We can hold about six units of information in working memory, and that information degrades very quickly. As a result, readers can focus on the meaning of a passage only if decoding and word identification processes are automatic. As text complexity increases, comprehension processes require more working memory. Once automatized, word identification uses little-to-no working memory, which frees it for use to monitor, integrate, and store the meaning of the text with other background knowledge.
The limits of working memory have a direct and immediate implication on benchmarks. They are set as progressive target levels of automatic passage reading to enable comprehension. Although automaticity often reaches 200-to-250 words correct per minute among literate adults, the initial goal is to achieve 150-to-200 words read correctly per minute by the spring of fourth grade. That seems to be the transition point from “learning to read” to “reading to learn.”
For students learning to read in Spanish, the fluency-comprehension association may be more complex. Word identification in Spanish is easier to achieve (and thus less taxing on working memory due to fewer sounds and a very consistent phonics pattern), while at the same time native Spanish-speaking ELLs may have relatively weak Spanish oral language skills. (Mancilla-Martinez, & Lesaux, 2011).
Alignment of English and Spanish Passage Reading Benchmarks
The limits of working memory and benefits of automaticity function similarly for both English and Spanish reading. As of fall 2015, the FastBridge research and development team began to adjust the FAST Spanish benchmarks to align better with the English standards. This process was completed for CBMreading in 2016. As of 2016, review and analysis were ongoing to consider the further alignment of earlyReading Spanish benchmarks with English, which set higher expectations for achievement.
Spanish Benchmarks are Distinct from Spanish Norms. As described in this document, the typical use of the term “benchmark” indicates broad reading proficiency among native Spanish speakers and the developmental progress toward requisite levels of automatic passage reading. The Spanish benchmarks do not align with the Spanish norms.
District or school managers who prefer alternate benchmarks can customize their benchmarks. For example, they may use the scores at the 15th and 40th percentiles from FAST norms. This would align the norms and benchmarks to the typical performance of (mostly) native English-speaking students in English-to-Spanish immersion settings. The FastBridge research and development team believes those benchmark standards are too low if the goal is Spanish-reading proficiency.
References
Christ, T. J. (2016). FastBridge Benchmarks and Norms Interpretation Guide. Minneapolis, MN: Developed by Theodore J. Christ and Colleagues. Distributed by FastBridge Learning.
Ehri, L. C. (2005). Learning to read words: Theory, findings, and issues. Scientific Studies of Reading, 9, 167-188.
Hoover, W. A., & Gough, P. B. (1990). The simple view of reading. Reading and writing, 2, 127-160.
LaBerge, D., & Samuels, S. J. (1974). Toward a theory of automatic information processing in reading. Cognitive Psychology, 6, 293- 323.
Mancilla-Martinez, J., & Lesaux, N. K. (2011). The Gap Between Spanish-speakers’ Word Reading and Word Knowledge: A Longitudinal Study. Child Development, 82, 1544–1560. http://doi.org/10.1111/j.1467-8624.2011.01633.x
Palmer, J., MacLeod, C. M., Hunt, E., & Davidson, J. E. (1985). Information processing correlates of reading. Journal of Memory and Language, 24, 59-88.
Savage, R. (2001). The 'simple view' of reading: Some evidence and possible implications. Educational Psychology in Practice, 17, 17- 33. 1
Dr. Helman is a Professor of Literacy Development and Director of the Minnesota Center for Reading Research at the University of Minnesota. She is a researcher and published author of Spanish curricula and assessments.