EVEA Project Deliverable <strong>Use</strong> <strong>and</strong> <strong>Validity</strong> <strong>of</strong> HLS <strong>in</strong> ELPA Systemsdom<strong>in</strong>ant L2, an oral version <strong>of</strong> the “Enhanced” HLS can be piloted. This oral version could bedeliberately targeted to speech communities where parents are known from experience to frequentlynot be literate <strong>in</strong> their L1 <strong>and</strong> largely non-fluent <strong>in</strong> <strong>English</strong> (e.g., families with recent refugee status).Moreover, with utilization <strong>of</strong> an oral version <strong>of</strong> an “Enhanced” HLS, the services <strong>of</strong> an <strong>in</strong>terpreter can be<strong>of</strong>fered to families (<strong>in</strong>terpreters are already recommended by some states like Vermont). Validationstudies can be carried out with the oral language adm<strong>in</strong>istration <strong>of</strong> the “Enhanced” HLS <strong>in</strong> wayscomparable to those described above for Study Design 3.Study Design 5: <strong>Use</strong> <strong>of</strong> additional or alternate measures <strong>of</strong> student language backgroundAbedi (2008) <strong>and</strong> Gonzalez et al., (1996) argue the need for multiple measures <strong>of</strong> student languagebackground rather than reliance on an HLS alone. <strong>The</strong>refore, we suggest that states consider the<strong>in</strong>clusion <strong>of</strong> additional measures <strong>of</strong> student language background to determ<strong>in</strong>e if the use <strong>of</strong> comb<strong>in</strong>edmeasures <strong>in</strong>creases their current hit rates (i.e., decreas<strong>in</strong>g the number <strong>of</strong> students who exit <strong>English</strong>language services at the immediate next ELPA adm<strong>in</strong>istration or who need referral to <strong>English</strong> languageservices by classroom teachers <strong>in</strong> the months immediately follow<strong>in</strong>g the HLS). For example, <strong>in</strong>formationabout k<strong>in</strong>dergarten <strong>and</strong> first grade students’ preschool experiences might be considered, particularlys<strong>in</strong>ce the number <strong>of</strong> EL students is vastly grow<strong>in</strong>g among the preschool population (Mathews & Ewen,2006) <strong>and</strong> most K-12 EL students are <strong>in</strong> the very earliest grades. Where students have attendedpreschool, data might be gathered <strong>in</strong> the form <strong>of</strong> a separate questionnaire completed by preschool staff<strong>and</strong> can be based on their first-h<strong>and</strong> observations <strong>and</strong> rat<strong>in</strong>gs <strong>of</strong> such factors as (1) opportunities forstudent exposure to <strong>and</strong> usage <strong>of</strong> <strong>English</strong> <strong>and</strong> their L1, <strong>and</strong> (2) the degree <strong>of</strong> student engagement with<strong>English</strong>-speak<strong>in</strong>g students <strong>and</strong> adults. In addition, preschool staff could provide <strong>in</strong>formal rat<strong>in</strong>gs <strong>of</strong><strong>English</strong> language pr<strong>of</strong>iciency (for example observation <strong>and</strong> rat<strong>in</strong>g protocols see Bailey, Huang, Osipova,& Beauregard, 2010). <strong>The</strong> validation <strong>of</strong> these additional measures alongside either the state’s st<strong>and</strong>ardHLS or an “Enhanced” HLS can also be carried out much like that described <strong>in</strong> Study Design 2 above.Unfortunately, implement<strong>in</strong>g multiple preschool language observations <strong>and</strong> assessments to pass on tok<strong>in</strong>dergarten staff may currently be outside the scope <strong>of</strong> the many private preschool programs that donot receive state or federal fund<strong>in</strong>g <strong>and</strong> are not m<strong>and</strong>ated to assess. Also hav<strong>in</strong>g school districtsadm<strong>in</strong>ister multiple assessments or <strong>in</strong>terviews with newly enrolled students <strong>in</strong> the higher grades is alsolikely to be impractical given current limitations placed on school budgets <strong>and</strong> staff<strong>in</strong>g. We recognizethat despite compell<strong>in</strong>g reasons to <strong>in</strong>clude multiple measures such as those made by Abedi (2008) <strong>and</strong>Gonzalez et al. (1996), there rema<strong>in</strong>s a strong impetus for a s<strong>in</strong>gle measure to <strong>in</strong>itially determ<strong>in</strong>e thelanguage dom<strong>in</strong>ance <strong>of</strong> students.<strong>The</strong>refore, we hark back to the beg<strong>in</strong>n<strong>in</strong>g <strong>of</strong> this paper <strong>and</strong> aga<strong>in</strong> po<strong>in</strong>t out that federal regulations aswell as the OCR memor<strong>and</strong>um <strong>and</strong> policy do NOT m<strong>and</strong>ate state use <strong>of</strong> an HLS for <strong>in</strong>itial identification <strong>of</strong>students even as almost all states do m<strong>and</strong>ate such a policy for their districts. One read<strong>in</strong>g <strong>of</strong> thedisparate practices around the nation <strong>and</strong> the conflicts <strong>in</strong> the research f<strong>in</strong>d<strong>in</strong>gs suggests an alternativestudy design <strong>in</strong>volv<strong>in</strong>g the ab<strong>and</strong>onment <strong>of</strong> the HLS entirely <strong>and</strong> try<strong>in</strong>g out <strong>in</strong> its stead a short, <strong>in</strong>cisivescreen<strong>in</strong>g tool for all <strong>in</strong>-com<strong>in</strong>g students.<strong>The</strong> field <strong>of</strong> education should be <strong>in</strong> a position to now figure out a way to quickly target some keylanguage competencies for such an <strong>in</strong>itial screener. Specifically, the screener should <strong>in</strong>clude ageappropriate oral language skills <strong>and</strong> possibly both oral language <strong>and</strong> literacy skills for older students firstenroll<strong>in</strong>g <strong>in</strong> U.S. schools <strong>in</strong> the higher grades—skills that have been found to be predictive <strong>of</strong> laterlanguage pr<strong>of</strong>iciency <strong>and</strong>/or academic success (e.g., recognition <strong>of</strong> derived word forms, ability to give18
EVEA Project Deliverable <strong>Use</strong> <strong>and</strong> <strong>Validity</strong> <strong>of</strong> HLS <strong>in</strong> ELPA Systemsformal def<strong>in</strong>itions). Such a screener should capitalize on digital technology to make the assessmentquick <strong>and</strong> uniform to adm<strong>in</strong>ister (e.g., computer-based or h<strong>and</strong>-held devices for teachers <strong>and</strong> olderstudents) <strong>and</strong> the results immediately available to a data management system (e.g., a web-based, datadrivensystem). <strong>The</strong> validation plan described <strong>in</strong> Study Design 1 or even an experimental design withr<strong>and</strong>omization <strong>of</strong> families to HLS or the <strong>in</strong>itial screen<strong>in</strong>g tool can be used to determ<strong>in</strong>e if hit rates differ(all th<strong>in</strong>gs be<strong>in</strong>g equal) across the two <strong>in</strong>struments, thus test<strong>in</strong>g the comparative efficacy <strong>of</strong> such auniversal screen<strong>in</strong>g tool. 8Recommendations <strong>and</strong> Conclusions<strong>The</strong> focus <strong>of</strong> this review has been primarily on the question phras<strong>in</strong>g <strong>and</strong> content, <strong>and</strong> theimplementation <strong>and</strong> <strong>in</strong>terpretation <strong>of</strong> exist<strong>in</strong>g HLS across EVEA states <strong>and</strong> around the nation that mayimpact the validity <strong>and</strong> reliability <strong>of</strong> the <strong>in</strong>formation yielded. However, we wish to stress that the localadm<strong>in</strong>istration <strong>of</strong> an HLS is another critical area that needs careful attention <strong>and</strong> <strong>in</strong>vestigation by statesas adm<strong>in</strong>istration practices also impact the quality <strong>of</strong> the validity <strong>of</strong> the <strong>in</strong>formation garnered with anHLS. Anecdotal <strong>in</strong>formation about local practices <strong>in</strong> HLS adm<strong>in</strong>istration suggests the completion <strong>of</strong> theHLS can easily be open for abuse with, for <strong>in</strong>stance, reports <strong>of</strong> school staff respond<strong>in</strong>g to the surveys forfamilies they th<strong>in</strong>k speak little or no <strong>English</strong> rather than allow<strong>in</strong>g families the chance to complete thesurvey themselves (e.g., provid<strong>in</strong>g translated versions <strong>of</strong> the HLS).We have several recommendations for improvement <strong>in</strong> current state practices with measures <strong>of</strong> thehome language construct, <strong>in</strong>clud<strong>in</strong>g two areas <strong>of</strong> recommendation at the level <strong>of</strong> the federalgovernment, <strong>and</strong> two areas <strong>of</strong> recommendation at the state level.Federal-Level RecommendationsRecommendation One: Provid<strong>in</strong>g state guidance<strong>The</strong> U.S. Department <strong>of</strong> Education must provide greater guidance on the use <strong>of</strong> HLS <strong>in</strong>clud<strong>in</strong>g thepossibility <strong>of</strong> encourag<strong>in</strong>g a universal screen<strong>in</strong>g tool as a viable alternative. By fund<strong>in</strong>g or conduct<strong>in</strong>g amore comprehensive analysis <strong>of</strong> state practices than the one we were able to conduct here, a reportfrom the Department can <strong>of</strong>fer concrete <strong>in</strong>formation about the k<strong>in</strong>ds <strong>of</strong> HLS question phras<strong>in</strong>g <strong>and</strong>solicited content that states should avoid <strong>in</strong> state-created HLS or <strong>in</strong> their own guidance to schooldistricts. From our own limited analyses, we predict that survey items that are ambiguously worded orfocus on a child’s first-acquired language will fail to yield pert<strong>in</strong>ent <strong>in</strong>formation on student’s currentlanguage dom<strong>in</strong>ance <strong>and</strong> exposure, two areas that are important for current language pr<strong>of</strong>iciency. <strong>The</strong>Department can also provide <strong>in</strong>formation about reported best practices across states for the proceduresused <strong>in</strong> survey implementation <strong>and</strong> <strong>in</strong>terpretation (<strong>in</strong>clud<strong>in</strong>g how to judge <strong>and</strong>/or give weight to certa<strong>in</strong>question responses). F<strong>in</strong>ally, the Department can dissem<strong>in</strong>ate validation studies that have been8 With a universal screen<strong>in</strong>g tool, native-<strong>English</strong>-speak<strong>in</strong>g students may also be identified as need<strong>in</strong>g furtherassessment for <strong>English</strong> language services. It is outside the scope <strong>of</strong> the current paper to discuss the possibility thatthese cases may not be false positives but rather students with genu<strong>in</strong>e language needs, particularly <strong>in</strong> the area <strong>of</strong>academic uses <strong>of</strong> language. At the very least, a universal screen<strong>in</strong>g tool as described here may prove useful <strong>in</strong>k<strong>in</strong>dergarten by also serv<strong>in</strong>g as an early <strong>in</strong>dicator <strong>of</strong> the need for further speech-language-hear<strong>in</strong>g evaluation. Thisadditional purpose would <strong>of</strong> course necessitate its own validity argument <strong>and</strong> validation plan.19