22  Reliability of ROAR-Fonema

22.1 Background: Published studies

Bhat et al. (2024) reported marginal reliability of 0.85 for ROAR-Fonema (see more detail in Section 20.1).

22.2 Reliability of fixed-length ROAR-Fonema

Table 22.1 reports marginal reliability computed based on data from 4129 students under the IRT model for 2 different versions of ROAR-Fonema: the short, 20 item version and the long 54 item version. The computer-adaptive ROAR-Fonema is planned for release in fall 2024 and will be more efficient and reliable with fewer items. Reliability (\(\rho_{xx^\prime}\)) is computed based on the estimated variance of \(\hat{\theta}\) relative to the estimated standard error (\(\widehat{SE}(\hat{\theta})^2\)) using Equation 22.1:

\[ \hat{\rho}_{xx^\prime} = \frac{\widehat{VAR}(\hat{\theta})}{\widehat{VAR}(\hat{\theta}) + \widehat{SE}(\hat{\theta})^2}, \tag{22.1}\]

Version Grade Empirical Reliability N
All All 0.82 4129
long 1 0.52 513
long 2 0.72 578
long 3 0.78 561
long 4 0.81 519
long 5 0.87 577
long 6 0.88 592
long K 0.35 389
short 1 0.6 62
short 2 0.74 132
short K 0.51 206
Table 22.1: Reliability of ROAR-Fonema by Grade

To ensure that ROAR-Fonema is fair and equitable for different demographic groups, we also report reliability by gender (Table 22.2), eligibility for free and reduced price lunch (Table 22.3), English learner status based on state of California designations (Table 22.4), primary langauge spoken (Table 22.5), special education (Table 22.6), ethnicity (Table 22.7), and race (Table 22.8).

Version Gender Empirical Reliability N
All All 0.81 3518
long F 0.83 1484
long M 0.82 1640
short F 0.67 178
short M 0.67 216
Table 22.2: Reliability of ROAR-Fonema by Gender
Free/Reduced Lunch Status Empirical Reliability N
All 0.67 394
Free 0.59 191
Paid 0.71 139
Reduced 0.71 64
Table 22.3: Reliability of ROAR-Fonema by FRL (California Sub-sample Only)
English Learner Status Empirical Reliability N
All 0.67 394
English Learner 0.58 257
English Only 0.73 100
Initial Fluent English Proficient 0.75 28
Reclassified Fluent English Proficient 0.77 9
Table 22.4: Reliability of ROAR-Fonema by EL Status (California Sub-sample Only)
Primary Language Empirical Reliability N
All 0.67 361
English 0.72 157
Spanish 0.59 204
Table 22.5: Reliability of ROAR-Fonema by Primary Language (California Sub-sample Only)
Special Education Status Empirical Reliability N
All 0.67 394
No 0.68 358
Yes 0.48 36
Table 22.6: Reliability of ROAR-Fonema by Special Education Status (California Sub-sample Only)
Hispanic Ethnicity Empirical Reliability N
All 0.63 375
No 0.54 28
Yes 0.63 347
Table 22.7: Reliability of ROAR-Fonema by Hispanic Ethnicity (California Sub-sample Only)
Race Empirical Reliability N
All 0.63 375
Hispanic 0.63 347
White 0.55 26
Table 22.8: Reliability of ROAR-Fonema by Race (California Sub-sample Only)

References

Bhat, Kruttika G., Alexa Mogan, Ana Saavedra, Mia Fuentes-Jimenez, Julian M. Siebert, Wanjing Anya Ma, Carrie Townley-Flores, et al. 2024. “Shared and Unique Influences of Phonological Processing on Reading and Math.” OSF Preprints. https://doi.org/https://doi.org/10.31219/osf.io/em3bg.