Research “I Will Study More . . . and Pray” Metacognition About High-Stakes Test Preparation Among Culturally & Linguistically Diverse Students Debra A. Giambo Introduction English at home, including both ELLs The CAHSEE graduation requirement and students fluent in English (California was temporarily suspended in January High-stakes testing continues in the Department of Education, 2015c). In 2013- 2016 for graduating classes between 2015 U.S. public schools despite a plethora of 2014, 17% (900,476) of students enrolled and 2018 with a grandfather clause for concerns from the field (e.g., Council of the in Texas’ public school were ELLs (Texas students that went back to 2003. Great City Schools, 2015; Krashen, 2012a, Education Agency, 2014). In 2015-2016, From 1986 to 2011, the Texas Education 2012b; Lazarín, 2014; Ravitch, 2011, 2016a, 9.8% (273,570) of Florida’s public school Agency required the passing of a high- 2016b) and among parents nationwide students were classified as ELLs (Florida stakes test for high school graduation and, (e.g., Brown, 2015; Layton, 2015; Wallace, Education Department, 2016a). subsequently, end-of-course assessments 2015). While there have been some modi- In these same states, scores from the in 2012-2014. Requirements have been fications to testing requirements in some 2015 National Assessment of Education revised for students who are 11th or 12th states (e.g., Layton, 2015), the majority of Progress (National Center for Educa- graders in 2014-2017 to stipulate that a students and teachers in the U.S. continue tion Statistics, 2016) demonstrate a gap student who has failed no more than two to experience a rigorous emphasis on test between the reading proficiency of 4th end-of-course assessments may earn a preparation. and 8th grade ELLs and non-ELLs. In high school diploma by committee decision For culturally and linguistically diverse California, 28% of 4th grade ELLs demon- (Texas Education Agency, 2016a, 2016b). students, some of whom have not yet strated proficiency in reading, compared Florida implemented the Florida achieved academic English proficiency, to 72% of students who were not ELLs. Comprehensive Assessment Test (FCAT), high-stakes testing can bring additional Twenty-two percent of 4th grade ELLs in a statewide assessment of students’ issues and challenges (Coltrane, 2003; Texas demonstrated proficiency in reading, academic content, in 1996. In 1999, the Hopkins et al., 2013; Huang, Han, & whereas 78% of non-ELLs demonstrated legislature added the use of the FCAT for Schnapp, 2012; Luykx et al., 2007, Mahon, reading proficiency. In Florida, 9% of 4th graduation from high school and a grading 2006; Murphy, 2007; Robinson-Cimpian, grade ELLs’ scores indicated proficiency in system for schools based on FCAT scores. Thompson, & Umansky 2016; Sanchez et reading, compared to 91% of students who Subsequently, in 2001, the State Board al., 2009; Solórzano, 2008). For English lan- were not ELLs. of Education established passing scores guage learners (ELLs), high-stakes test- For 8th grade ELLs and non-ELLs in on the FCAT to determine eligibility for ing in English becomes a test of English each of these states, the gap in reading a standard high school diploma. As of the proficiency, rather than a test of content proficiency scores was even wider: In Cal- 2015-2016 academic year, Florida contin- knowledge (Tsang, Katz, & Stack, 2008), ifornia, the gap, respectively, was 14% and ues to require students to pass the reading/ therefore raising questions regarding the 86%; the scores in Texas, respectively, were English language arts subtest of high- reliability and validity of such assessments 11% and 89%, and in Florida, they were 5% stakes standardized test for graduation, for students with limited English proficien- and 95% (National Center for Education now the Florida Standards Assessment cy (Hopkins et al, 2013; Robinson-Cimpian Statistics, 2016). (FSA), and also requires passing an end et al., 2016). As of 2011-2012, California, Texas, and of course assessment in Algebra I (Florida California, Texas, and Florida had the Florida each required the passing of at Department of Education, 2015b). highest number of language minority stu- least the reading/language arts and math- In California, Texas, and Florida, the dents among U.S. states as of 2012-2013 ematics subtests for high school gradua- reported passing rates on the high-stakes (National Center for Education Statistics, tion (Center on Education Policy, 2012; test required for graduation indicate that 2015a). In California, in the fall of 2014, National Center for Education Statistics, White, non-Hispanic students showed the 22.3% (1,392,263) of the students in public 2011-12), although such requirements highest passing rate when compared to all schools were ELLs. Additionally, 42.9% have been in transition in recent years. students, Hispanic students, and ELLs. (2,672,128) speak a language other than The California High School Exit Exam- Significant numbers of Hispanic students ination (CAHSEE) was a requirement for and ELLs in Texas and Florida have not Debra A. Giambo is a professor earning a high school diploma beginning passed the assessments. Results reported in the Department of Curriculum, Instruction, with the class of 2006 (California De- here include selections of the most recently and Culture in the College of Education partment of Education, 2015a) and was reported passing rates gleaned from each at Florida Gulf Coast University, preceded by a comprehensive competency of the states’ education agencies’ websites. Fort Myers, Florida. exam (Center on Education Policy, 2009). In California, 85% of all students are MULTICULTURAL EDUCATION 26 Research reported to have scored as passing the (Florida Department of Education, 2015a). lies with the examination of culturally and CAHSEE, while 93% of White, non-His- However, it is possible that the actual linguistically diverse students’ perceptions panic students scored as passing, 80% drop-out rate was considerably higher of a high-stakes test and their self-reflection of Hispanic students, and 43% of ELLs due, in part, to the withdrawal codes used on how they prepare for and adjust to the passed the exam (California Department to indicate that a student has left school. requirements of confronting the test. of Education, 2015b). Passing rates in Among several withdrawal codes, only one In other words, self-perceptions of ef- Texas on EOC exams for English II (as an indicated that a student has dropped out fectiveness in test preparation influences example) in 2014-2015, indicate that 44% of school. Other codes could be used when actions so that students who engage in of White, non-Hispanic students scored as a student leaves without explanation or to self-regulatory processes are more likely passing, while 35% of all students, 34% enter a GED program, for example, and the to adapt and change. Thus, guided by so- of Hispanic students, and 18% of ELLs student would not be counted as dropping cial cognitive theory, it can be anticipated scored as passing (Texas Education Agen- out of school, even if the student drops that students who are self-reflective in cy, 2015). out of the GED program (Giambo, 2009; their preparation for a high-stakes test In Florida, passing rates on the FCAT Merrow, 2004). may be more likely to make changes in 2.0, required for a standard high school A separate code indicated that a stu- their preparation or subsequent academic diploma in 2013 and 2014, were consis- dent’s whereabouts was unknown (Florida behavior. tent across these two years for a variety Department of Education, 2015a). Consid- Student perceptions of assessments of student groups in 10th grade (Florida ering the high-stakes testing requirement affect the manner in which they approach Department of Education, 2016b). In 2014, for high school graduation, the ELL pass preparation as well as test results, and statewide assessment results indicated rate, the use of withdrawal codes, and the this has been found with middle and that 55% of all students passed. When sep- pressure on school districts to show con- high school students as well as university arating out White, non-Hispanic students, tinual improvement, it is possible that the students. When students perceive an as- and Hispanic students, scores indicated drop-out rate was under-reported and that sessment as inappropriate or inauthentic, 68% passing and 50% passing, respectively. the graduation rate, especially for ELLs, they employ more superficial methods Results for ELLs in Florida showed a was over-reported (Giambo, 2009). of preparation than if they perceive the passing rate of 11%. Scores for 2013 saw assessment as valuable and appropriate similar patterns with a difference of 1-2% Investigating Perceptions (Alkharusi, 2013; Hong & Peng, 2008; lower passing rates for each of the student Lizzo, Wilson, & Simons, 2002). Increased Since high stakes testing can be an groups mentioned above. While the passing preparation efforts for a test perceived as inaccurate measure of educational achieve- rates for ELLs and Hispanic students are valuable, however, do not necessarily mean ment for culturally and linguistically lower than that of other student groups that students who perceive themselves diverse students (Coltrane, 2003; Hopkins discussed, overall, it is important to note as using metacognitive strategies will do et al., 2013; Huang et al, 2012; Luykx et that, depending on the reporting forum, better on the test (Hong & Peng, 2008; al., 2007, Mahon, 2006; Murphy, 2007, Rob- high-stakes test passing rates of ELLs in Purpura, 1997). inson-Cimpian et al., 2016; Sanchez et al., Florida may have been inflated in the past Students who anticipate poor perfor- 2009; Solórzano, 2008; Tsang et al., 2008), (Giambo, 2009). mance on a test are more likely to attribute an important step toward more promising Graduation rates in California, Texas, their score to external factors, such as dis- opportunities for this group of students is and Florida indicate patterns similar to tractions in the testing situation as well as to investigate their perceptions of the test the passing rates of the high-stakes tests in issues or situations outside of the test, than and effective test preparation as well as which we see discrepancies between White, those who expect a good outcome (Gaier, the effects on their motivation to achieve non-Hispanic students, Hispanic students, 1962). Increased efficacy in approaching academically. and ELLs. Data reported by the National the goal of effectively helping culturally Bandura’s social cognitive theory in- Center for Education Statistics (2015b) and linguistically diverse students im- cludes the perspective that perceptions of indicated that graduation rates for ELLs prove performance on high-stakes assess- self-efficacy influence motivation and action and Hispanics remained below those of ments may involve better understanding and are affected by social factors (Bandura, White, non-Hispanic students, and all stu- of their approaches to testing via their 1991). Central to this theory is the role of dents combined in these states. Calculated perceptions of: the test, their preparation “cognitive, vicarious, self-reflective, and using the Adjusted Cohort Graduation Rate for the test, and the effect of the test on self-regulatory processes” in human ad- (ACGR), the 2013-2014 graduation rates for academic behaviors. aptation and change (Bandura, 1989). The all students, White, Hispanic, and ELLs, are The purpose of this study was to inves- relevance of Bandura’s theory to this study displayed in Table 1. tigate the following research questions: In addition to considering the high- stakes-test passing rates, an understand- ing of the drop-out rates can provide a Table 1 more complete picture of, for example, Adjusted Cohort Graduation Rate (2013-2014) for students in California, Texas, and Florida Florida’s ELLs’ academic achievements States All students White, Hispanic ELLs in high school. For example, the Florida non-Hispanic students Department of Education reports that, in students 2006, the statewide drop-out rate in Flor- California 81% 88% 77% 65% ida was 3.5% (2008), and, in 2013-2014, Texas 88.3% 93% 85.5% 71.5% the drop-out for all students fell to 1.9% Florida 76.1% 81.7% 75% 55.8% and was 2.3% for Hispanic male students SPRING/SUMMER 2017 27 Research 1. What do culturally and linguistically more in-depth information. Consideration spoke Haitian Creole, and one spoke both diverse students think and feel about was given to the possibility that students English and Haitian Creole equally fre- high-stakes testing (i.e., FCAT)? completing the survey may have some quently (with one non-response). 2. What are culturally and linguistically limitations in their English proficiency as Of the 38 participants in the focus group diverse students’ perceptions regarding well as the possibility of limited motivation interviews, 23 were middle school students effective preparation for high-stakes to complete the survey, so the focus group and 15 were high school students. All par- testing (in classes and individually)? interview was intended to be an oppor- ticipants were from schools in two adjacent tunity for students to provide additional county school districts. Of the 23 middle 3. Do culturally and linguistically diverse students perceive that high-stakes information. school focus group participants, 14 report- testing affects their academic motivation, ed that they had received ELL services in resulting in perceived differences in Participants the past, and nine reported never having academic behavior? received ELL services. Nine of the 15 high Participants in both the survey and school focus group interview participants the focus group interviews were cultural- Methodology ly and/or linguistically diverse students reported that they were currently receiving ELL services, one had received services in Measures in South Florida. Survey participants the past, and five reported never having included elementary (4th to 6th graders) Measures employed in this study received ELL services. and middle and high school students, and included a survey and a focus group in- Focus group participants also report- focus group participants included middle terview protocol. Both the survey and the ed on their use of languages other than and high school students. All were active in interview protocol were developed in con- English. Among the middle school partic- programs that support underrepresented junction with three high school ELL teach- ipants, six spoke Spanish, 14 spoke Hai- students in their preparation for post-sec- ers, each with 15 to 26 years experience in tian Creole (seven of these also reported ondary education, offered through local col- education, and with an interdisciplinary speaking French), and the remaining three leges and universities in cooperation with panel of researchers. spoke Arabic, Tagalog, or Hungarian. The the students’ home schools. Participation The survey questions were examined by high school participants reported a similar was limited to those for whom permission these stakeholders for balance in terms of distribution, with three who spoke Span- to participate had been obtained. allowing for both positive and negative re- ish, 11 who spoke Haitian Creole (four of All participants had diverse back- sponses, appropriateness, responsiveness whom also reported speaking French), and grounds based on ethnicity (i.e., their to research questions, balance between one who spoke French and Dutch. families had come to the U.S. from another Likert-type and free responses, and level country) and/or language (i.e., they were of difficulty in terms of possible limitations experiencing or had experienced limited Procedures in participants’ English proficiency. English proficiency). Many participants Survey The survey included 18 items divided had come to the United States at a variety into three parts. Part one included four To maximize participants’ comfort level, of ages while some had been born in the open-ended questions and one yes/no item the surveys were administered by staff U.S. At the time of the interviews and sur- focusing on students’ future plans; part members of the college-bound programs, veys, all participants demonstrated facility two was composed of six items requiring and assistance was available, if needed, with English sufficient for completion of Likert-type scale responses that focused on during the completion of the survey. Partic- the survey and/or focus group interview students’ general perceptions of the FCAT; ipants completed the surveys during one of based on consultations between and rec- and part three included five open-ended their college or university on-campus visits. ommendations from the college program questions and two yes/no questions in- coordinator and the cooperating teacher tended to elicit more specific information Focus Group Interviews at the participants’ schools. on students’ perceptions. The 62 survey participants included Participants were interviewed in the Likert scales are often viewed as an 25 elementary students, 12 middle school college-bound program offices during one effective way to measure attitudes (Al- students, and 25 high school students. Ac- of their on-campus visits. Most partici- cázar-Olán, Deffenbacher, Escamilla-Tecal- cording to self-report, nine were currently pants had been to these locations multi- co, 2016; Geldhof, Gestsdottir, Stefansson, receiving ELL services, nine had received ple times, so the location was familiar to 2015; Likert, 1932; Page-Bucci, 2003) and ELL services in the past, and 44 had never them. The interviews were conducted by a seem to be similar in reliability and valid- received ELL services. Additionally, sur- researcher and a graduate student, lasted ity to traditional measures of self-efficacy vey participants self-reported information approximately 45 minutes, and included an (Maurer & Andrews, 2000). Additionally, the about their first language, home language, average of six participants, with a range use of focus group interviews to elicit mean- and language used with friends. Thirty-sev- from four to 10, depending on logistics. ingful information on perspectives has long en reported speaking English as their first An interview protocol was followed been recognized as effective (Butler, 2002; language, seven spoke Spanish as their first in each interview which allowed for in- Center for Disease Control, 2008; De Groot, language, and 17 spoke Haitian-Creole as troductions followed by an emphasis on 2002; Israel & Galindo-Gonzalez, 2014; their first language, with one unreported. the importance of each person’s opinion. Patrick & Middleton, 2002; Zhu, 2005). At home, 40 spoke English, nine spoke Participants were told the purpose of The focus group interview protocol Spanish, and 13 spoke Haitian Creole, and the interviews (i.e., to understand their consisted of five open-ended questions this includes one participant who reported thoughts about the FCAT) and logistical with supporting prompts. The content speaking both English and Haitian Creole rules (i.e., allowing each person to respond was intentionally similar to the categories equally frequently at home. With friends, without others talking, not needing to of topics on the survey in order to elicit 55 used English, two spoke Spanish, three agree with anyone else, respecting others MULTICULTURAL EDUCATION 28 Research and their responses, and respecting the stating that the FCAT involves unrealistic Research Question 2: confidentiality of responses following the standards, wastes time, and provides pres- What are culturally and linguistically close of the interview). sure for students newly arrived to the U.S. diverse students’ perceptions regarding Handwritten notes were taken during One view, expressed by a middle school effective preparation for high-stakes the interviews, but interviews were not participant, was reflective of a frequently testing (in classes and individually)? audio-taped to encourage participation expressed concern: “It can be a bad day and Responses to relevant survey ques- and reduce anxiety. Numbers were used for can affect the rest of your life.” tions were themed to examine trends in identification purposes both on the table The majority of the middle school and responses. In some instances, participants during the interview and in the notes to high school participants indicated that indicated more than one idea in response maintain anonymity in the written record they do not feel that the FCAT allows to a question, and, in those cases, the total of the interviews. At the conclusion of the them to show a lot of what they know and number of responses exceeds the number interviews, students were thanked and so may consider the test to be an unfair of participants. Responses were tallied reminded to respect the confidentiality of assessment of their knowledge and skills, among themes, and tallies are provided everyone’s responses. while the elementary participants again without percentages due to the occurrence responded more positively. Specifically, of multiple ideas from some participants. Data Analysis 68% (25 out of 37) of middle and high Regarding student perceptions of what school participant responses indicated that Data were entered into three databases: is helpful in doing well on the assessment, the FCAT either lets them show “some” one for Likert-type survey responses, one total responses indicated the following were to “not much” of what they learned in for survey free-response items, and one helpful, listed in descending frequencies: school. However, the majority of elemen- for focus group interview responses. Free studying (15), learning in class (12), teach- tary participants indicated that the FCAT responses were themed, and both numeri- ers (6), faith/confidence/positive thinking/ gave them an opportunity to show “a lot” cal and free response items were analyzed perseverance (5), and five did not provide or “more than some” of what they know: for frequency of responses. Direct quotes a response to the question (See Table 3.) 84% (21 out of 25) indicated showing a lot were entered into the database to enhance While some responses indicate an of what they know, 12% (three out of 25) and clarify responses and to maintain and awareness of the benefits of actively par- responded they can show some of what illustrate student voices. ticipating in preparing to do well on the they know on the FCAT. assessment (e.g., studying, learning in Results Table 2 Research Question 1: Survey Responses Regarding Feelings about the FCAT What do culturally and linguistically diverse students think and feel about Responses Elementary Middle High Totals high-stakes testing (i.e., FCAT)? n = 25 n = 12 n = 25 n = 62 While elementary participants in this 1 - Good 17 (68%) 0 1 (4%) 18 (29%) study tended overall to offer more positive 2 3 (12%) 1 (8%) 4 (16%) 7 (11%) responses about how the FCAT makes 3 - Neither good nor bad 5 (20%) 9 (75%) 12 (48%) 26 (42%) them feel, the majority of middle and high 4 0 1 (8%) 2 (8%) 3 (5%) school participant responses indicated a 5 - Bad 0 1 (8%) 6 (24%) 7 (11%) neutral view of the assessment. Among the elementary school participants, 68% (17 out of 25) responded on the survey that the Table 3 Participant Perceptions of What Helps Them Do Well on the High-Stakes Test FCAT makes them feel good, while 58% (21 (number of survey responses in order of frequency) out of 37) middle and high school students indicated that the FCAT made them feel What helps me do well Elementary Middle High Total number of responses neutral (i.e., neither good nor bad) (See Studying 5 5 5 15 Table 2 for specifics). Learning in class 5 2 5 12 Focus group interviews with middle The teacher 5 0 1 6 and high school students indicated that Faith, confidence, positive thinking, 39% (15 out of 38) of participants agreed perseverance 3 0 2 5 with a negative view of the FCAT. Students No response 1 0 4 5 indicated concern about the use of the Reading 2 0 2 4 Eating breakfast 0 3 1 4 FCAT to determine future educational Concentrating/ taking time 1 1 1 3 opportunities and graduation from high Re-reading passages and questions 1 1 0 2 school as well as having to take the FCAT Preparation 0 1 1 2 over and over beginning in 3rd grade. Music 0 1 1 2 A similar pattern of responses occurred Practicing FCAT-type questions 0 0 2 2 in the focus group interviews, with younger Sleeping enough the night before 1 1 0 2 students who were, in this case, middle Relaxing 2 0 0 2 school participants, demonstrating a more Using a dictionary 0 0 1 1 neutral view and high school participants Common sense 0 0 1 1 Knowing what to do 0 0 1 1 exhibiting stronger, negative reactions and Nothing 0 0 1 1 SPRING/SUMMER 2017 29 Research class, perseverance, reading, concentrat- prepare them as well (26%) while some of survey participants and nearly half of ing, taking time, re-reading questions and middle and high school participants felt focus group participants responded posi- passages, practicing FCAT-type questions, that teachers did not seem to care about tively. Survey data indicate that 56% (35 using a dictionary), these were unspecific their preparation or prepare them well out of 62 participants) thought the FCAT in nature, and many of the responses were (42%). There was some acknowledgment would help them in their classes, while further removed from active participation among the middle school participants that 37% (23 out of 62) responded negatively. in learning and improving preparation their language arts and mathematics class- Six percent provided no response. (e.g., faith, confidence, positive thinking, es were the only classes that were helpful Some examples of responses from the eating breakfast, preparation, music, in preparing for the assessment, as these focus group interviews include: “I see sleeping the night before, relaxing, com- were (at the time) the only subjects covered I don’t know stuff on the FCAT, [and I mon sense, knowing what to do, doing by the assessment. want to] do better next year”; “It makes nothing). Middle school participants focused more me want to think harder”; “I want to learn When asked what participants could on the dull colors used in the test as well as what they be talking about.” One student do to help themselves do well on the high- the irrelevance of the reading passages to explained that a friend described the stakes assessment, responses indicated the their lives (30%), while the high school par- FCAT as a video game: If you beat a level, following: studying (25), getting adequate ticipants concentrated more on the pres- you go on to the next level; your skills sleep (8), paying attention in class and on sure they felt regarding the high-stakes of improve as the levels get harder. Some the test (7), having breakfast (7), and seven the assessment (33%). While participants indicated this perception was due to the either did not respond or responded that brought up some issues and ideas that may lower pressure after the test is over. they did not know. There were between help students score somewhat higher on When asked how the FCAT helps or one and four responses indicating the fol- the assessment, comparison of responses in would not help them do better in their lowing would be helpful: staying calm (4), the focus groups and on the surveys shows classes, more than a third of participants practicing (3), studying the night before (2), consistency, overall, regarding the lack (34% or 21 out of 62) provided no response nothing (2), working hard (2), reading more of substance when considering effective (See Table 5.) Fifteen percent provided (2), understanding and trying to remember preparation for the assessment. responses within the theme that the FCAT (2), listening to music (1), reading the book shows them what they need to learn, while (1), trying one’s best (1), staying confident Research Question 3: 18% felt that the FCAT would not help in (1), not talking about the test (1), having Do culturally and linguistically their classes due to their view that they good attendance (1), doing nothing differ- diverse students perceive that high- did not see the relevance of the test to ent because of success on the test in the stakes testing affects their academic their classwork. Each of the additional past (1), preparing (without being specific; motivation, resulting in perceived responses were provided by 6% (four out of 1), and praying (1). (See Table 4). As with differences in academic behavior? 62) or fewer of participants. One motivat- the previous question, many responses ed participant pointed out that, although When asked if they thought they could indicated the benefits of actively partici- he wanted to learn from his mistakes on do better in their classes after taking the pating in preparation but were unspecific the test, the results were not provided in FCAT, and why or why not, more than half or not directly connected to preparation for itemized format, so he was unable to do so. a high-stakes test. When asked on the survey how classes Table 4 helped prepare participants for the FCAT, Perceptions on What Participants Can Do to Help Themselves Do Better on the Assessment 31% (19 out of 62) participants did not (number of survey responses in order of frequency) respond to the question. Fifteen responses (24%) indicated that classes help them What I can do to help me do better Elementary Middle High Total responses prepare without providing a response Study 12 4 9 25 as to how that occurs, 23% (14 out of 62) Sleep 4 3 1 8 indicated that they help because what Pay attention in class/on test 4 1 2 7 they learn is what they have to do on the Eat breakfast 5 1 1 7 FCAT. Additional responses were low in No response/ don’t know 1 1 5 7 frequency (i.e., teacher provides helpful Stay calm 3 0 1 4 tips for the test [2 responses] and teacher Practice 1 1 1 3 Study night before 0 0 2 2 enthusiasm [1 response]). Other responses Nothing 0 2 0 2 were not related to the question (i.e., enjoy- Work hard 1 1 0 2 ing classes, because classes help, and some Read more 0 0 2 2 questions on the FCAT are difficult). These Understand/try to remember 0 0 1 2 responses regarding classroom help in Listen to music 0 0 1 1 preparation for the FCAT were, therefore, Read the book 0 1 0 1 were predominantly either nonexistent or Try my best 1 0 0 1 nonspecific in nature. Stay confident 1 0 0 1 Not talk about the test 1 0 0 1 In the focus group interviews, partici- Have good attendance 1 0 0 1 pants varied on their responses about how Do nothing different (past success on test) 1 0 0 1 their classes helped to prepare them for Prepare 0 1 0 1 the assessment. Some middle school par- Pray 0 0 1 1 ticipants expressed that teachers helped MULTICULTURAL EDUCATION 30 Research Approximately one-third of focus group and others lamenting its insufficiencies Another stated, “It’s a useless test which participants responded that the FCAT (Cotterell, 2015; WESH, 2015), possibly does not properly evaluate students.” would not result in their doing better in due to the increased pressures resulting Many participants were clearly affected their classes, and that there seemed to be from the use of the test. by the pressure placed on them to pass a disconnect between the FCAT and class The use of such tests is particularly the test. One suggestion from participants work. For example, “[The] FCAT did not significant for culturally and linguistical- was to reduce this pressure, possibly by apply to class work”; “Because it’s not a ly diverse students. For these students, removing it as a graduation requirement. true test of what I learned in class.” Addi- high-stakes tests can pose significant and The issue of pressure was a recurring one tional remarks indicated that instructors specific challenges, over and above those in both the surveys and the focus group often do not want to discuss the FCAT after for many other students (Coltrane, 2003; interviews and one that seemed to color it is over: “If we ask a question in reading Hopkins et al., 2013; Huang et al, 2012; students’ views of the test. This pressure is [class] after the FCAT, they won’t explain Luykx et al., 2007, Mahon, 2006; Murphy, seen by some to be the result of changes in it and say to ask next year.” One student 2007, Robinson-Cimpian, 2016; Sanchez et the use of the results of the test, which has responded both positively and negatively al., 2009; Solórzano, 2008). Students from also resulted in high pressure for admin- and said that, although s/he gets nervous diverse backgrounds, who are also ELLs, istrators and teachers in addition to the on test day and forgets everything, s/he are particularly susceptible to these pres- students (ESOL teacher, name excluded recognized the benefit of having to focus sures, as high-stakes testing in English upon request, personal communication, to take the test. can become a test of English proficiency, March 20, 2015). rather than a test of content knowledge When asked to indicate what might Discussion (Tsang et al., 2008), thus rendering it im- be helpful in preparing effectively for the possible to meet the objectives of the test high-stakes assessment, including what The culturally and linguistically diverse (Hopkins et al, 2013; Robinson-Cimpian participants themselves can do to prepare, students in this study demonstrated a dif- et al., 2016). Culturally and linguistically responses to these items provided striking- ference in their view of the FCAT depend- diverse students, including ELLs, are even ly limited information, and the nonspecific, ing on their grade level. More elementary more vulnerable to testing pressures than superficial nature of many responses (e.g., participants had a positive view of the are many other students. “study more,” “sleep,” “classes help me assessment, while middle and high school While it is not surprising that middle learn”) as well as the lack of responses was participants’ views demonstrated neutral and high school students might view a meaningful. There was a plethora of non- or negative views of the assessment. Some required, time-consuming, high-stakes specific responses and a lack of responses middle and high school participants con- test negatively, the reasons for such a to these types of questions across questions nected this with the pressure to pass the view should be considered as significant. and measures, and when responses were test and concern over the use of the test Many participants were concerned about provided, they tended to be superficially scores. Thus, the results of this study in- the use of the test to determine graduation considered and non-specific. dicated a difference in viewpoint between from high school and future educational The relevance of high-stakes tests to younger and older students. opportunities. Many participants were also students’ learning and lives, academical- While the high-stakes test in Florida concerned that the test did not allow them ly and otherwise, was not clear for many has recently changed to the Florida Stan- to show what they knew and could do and participants. It is important to note that dards Assessment, the use of high-stakes that much of what was on the test had not according to previous research the rele- testing has largely remained the same. The been taught to them. One student said, vance a student assigns to a testing task FCAT was originally intended to measure “We should have a better way of testing.” can predict test performance (Roberts & student achievement of the Sunshine State Standards, but the use of the assessment has changed over time. For the years that Table 5 the participants in this study had been in Themed Perceptions of How the FCAT Helps/Does Not Help in Academic Classes school, the test was, in part, used as a gate- keeper, and students had to pass the test Participant survey responses n % to receive a standard high school diploma, No responses/don’t know 21 34% regardless of their grades in school. High-stakes tests in Florida have also Reasons why the FCAT helps The FCAT shows me what I need to learn. 9 15% been and continue to be used to assign I have to work harder and pay attention. 4 6% grades to schools, which can affect school There is less stress when the test is over. 3 5% funding from the state. Many parents, stu- The FCAT is a review of class content. 3 5% dents, and teachers in FL and nationwide The FCAT is preparation for the following year. 2 3% have protested an overuse of instructional The FCAT shows what I know. 2 3% time for the purpose of test preparation The FCAT gives teachers information (e.g., Brown, 2015; Jacobowitz, 2015; on class placement. 1 2% Shammas, 2016; WESH, 2015). In 2015, The FCAT gives me confidence. 1 2% a bill was passed in Florida to limit time Reasons why the FCAT does not help in class spent on standardized tests, and it received The FCAT does not apply to classwork. 11 18% mixed reviews from lawmakers, parents, The FCAT makes me feel like I have to know and teacher organizations, with some everything. 3 5% seeing it was a step in the right direction I already know the answers. 2 3% SPRING/SUMMER 2017 31 Research Dansereau, 2008), so these disconnects for Metacognitive skills can be increased be affected (Roberts & Dansereau, 2008). students may affect their performance on through classroom instruction (Tok, 2013; The initial, intended use of the FCAT as an the test as well as in their classes. Vrieling, Bastiaens, & Stijnen, 2012) as can assessment of student achievement of the The lack of responses to items asking critical thinking skills (Bangert-Drowns, & Sunshine State Standards, was obviously students to indicate the best way to pre- Bankert, 1990; Davoudi & Sadeghi, 2015; unclear to participants in this study. Help- pare for a high-stakes test as well as how Wismath, Orr, & Good, 2014). Metacogni- ing students to see connections between to effectively increase their achievement tive instruction has been shown to signifi- the content area standards, instruction on the test may be an indication of limita- cantly improve the ability of low-achieving and learning in class, and high-stakes tions in metacognitive skills and a lack of students’, including many of Hispanic or- testing may be a valuable use of time with awareness of diverse learning strategies. igin, to solve math problems (Cardelle-El- an eye to increasing test performance. More specifically, the superficiality as awar, 1992). There is some indication that Re-examination of the effects and the well as lack of responses on these items metacognitive skills develop along with effectiveness of employing high-stakes cannot be seen as indicative of interrupt- intellectual ability, but do not completely accountability measures in ways that de- ed completion of the survey, as the same depend on intellectual ability, and are a crease instructional time, that place more participants provided responses to subse- stronger predictor of learning performance of a burden on specific groups of students, quent items. It is possible that participants than is intelligence (Veenman, Wilhelm, and that fail to encourage and reinforce skipped these items due to limitations in & Beishuizen, 2004; Veenman & Spaans, critical thinking and connection-building their ability to know, process, and convey 2005). It is quite possible that spending ac- skills must be undertaken. Research evi- information about what they know. In ademic time on teaching students to think dence supporting such a re-examination other words, it is possible that participants about their own learning, rather than is readily available for the consideration lack metacognitive awareness to know and conducting a drill and practice type of test of policymakers. While accountability of to be able to explain what would be most preparation may be a more effective way students and teachers is politically pop- helpful to them both in preparation for a to help students raise test achievement. ular and can sometimes be educationally high-stakes test as well as in processing appropriate, accountability that is effective information about their learning and abil- Implications and fair, in which the objectives of the as- ities in relation to the test. sessment can be accomplished, and that and Recommendations Many participants reported spending enhances and improves instruction and time in class on test preparation involv- Responses from study participants, learning is both preferred and imperative. ing drill and practice and on test-based combined with the superficiality and lack workbooks. Although the research on ef- of specific responses in many cases, seem References fects of drill and practice test preparation to indicate that students may lack the for elementary, middle, and high school metacognitive skills to assess both their Alcázar-Olán, R. J., Deffenbacher, J. L., Esca- students is limited, the indications are preparation prior to testing as well as that milla-Tecalco, H. (2016). Developing a valid that such types of test preparation may which would help improve their prepara- version of an inventory to measure anger in not improve student test results. In one tion. It is possible that, with the pressure Mexican adolescents of middle school level: study, comparisons among test preparation associated with the use of high-stakes test The ML-STAXI-MS. Youth & Society, 48(1), procedures for 3rd and 5th grade students scores for school administrators, teacher, 126-143. Alkharusi, H. (2013). Canonical correlational found that teaching-to-the-test-style and students, the focus on test preparation models of students’ perceptions of assess- preparation did not improve students test drill may be taking time away from instruc- ment tasks, motivational orientations, and performance over teaching content guided tion that would enhance critical thinking learning strategies. International Journal by state standards (Welsh, Eastwood, & skills, even while such instruction can of Instruction, 6(1), 1308-1470. D’Agostino, 2014). also help improve academic performance Bangert-Drowns, R. L., & Bankert, E. (1990, For older learners, completion of prac- (Bangert-Drowns, & Bankert, 1990; Car- April). Meta-analysis of effects of explicit in- tice tests and tutoring for college-level delle-Elawar, 1992; Davoudi & Sadeghi, struction for critical thinking. Paper present- admission tests have been shown to have 2015; Tok, 2013; Veenman & Spaans, 2005; ed at the Annual Meeting of the American Educational Research Association, Boston, little to no affect on test scores (Briggs, Vrieling et al., 2012; Wismath et al., 2014) MA. ERIC ID# ED328614. 2007, 2009; Scholes & Lain, 1997). Drill and thus could, more directly, improve Bandura, A. (1989). Human agency in social and practice methods for university high-stakes testing scores. cognitive theory. American Psychologist, 44, students may be more effective when This potential is particularly important 1175-1184. students are engaged with the material for students who are more challenged by Bandura, A. (1991). Social cognitive theory of (Kamarulzaman & Shaari, 2015). Howev- a high-stakes test, such as culturally and self-regulation. Organizational Behavior er, “deliberate practice” has been shown linguistically diverse students. While this and Human Decision Processes, 50, 248-287. to be insufficient alone and affected by study addresses culturally and linguisti- Briggs, D. C. (2007). The effects of admissions test preparation: Evidence from NELS: 88. individual factors in improving certain cally diverse students, the results may not Retrieved from http://epicpolicy.org/files/ skills (Campitelli & Gobet, 2011). While be exclusive to them. More research in this Briggs_Theeffectofadmissionstestprepara- these drill and practice types of activi- area is needed, especially specific to diverse tion.pdf ties may help students to become more and ELL students. Briggs, D. (2009). Preparation for college admis- familiar with the format of the test, they Assisting students to make sense of the sion exams. 2009 NACAC Discussion Paper. may also take time away from instruc- relevance of high-stakes testing to their National Association for College Admission tion that enhances critical thinking and lives may also help increase test perfor- Counseling. Retrieved from http://www.nac- problem-solving skills while supporting mance. When there is such a disconnect acnet.org/PublicationsResources/Research/ Documents/TestPrepDiscussionPaper.pdf student learning regarding the standards. for students, their test performance may Brown, E. (2015, March 7). Some parents across MULTICULTURAL EDUCATION 32 Research the country are revolting against standard- viewing: Students and teachers talk about Jacobowitz, R., & Tobin, K. T. (2015). Time on ized testing. The Washington Post. Retrieved learning and schooling. Educational Psychol- test: The fixed costs of 3-8 standardized test- from https://www.washingtonpost.com/local/ ogist, 37(1), 41-52. ing in New York State. New Paltz, NY: State education/some-parents-across-the-country- Florida Department of Education. (2008). University of New York at New Paltz: The are-revolting-against-standardized-test- Graduation and dropout rates by district, Benjamin Center for Public Policy Initiatives. ing/2015/03/05/e2abd062-c1e1-11e4-9ec2- 2005-06 school year. Education Information Retrieved from http://www.newpaltz.edu/ b418f57a4a99_story.html & Accountability Services, Data Publications media/the-benjamin-center/Time%20on%20 Butler, D. L. (2002). Qualitative approaches & Reports. Retrieved from http://www.fldoe. Test%20-%20The%20Fixed%20Costs%20 to investigating self-regulated learning: org/eias/eiaspubs/2006grad.asp of%203-8%20Standardized%20Testing%20 contributions and challenges. Educational Florida Department of Education. (2009). in%20New%20York%20State.pdf Psychologist, 37, 1, 59-63. Student performance results: Demographic Kamarulzaman, M. S., & Shaari, A. J. (2015). California Department of Education. (2015). report. Retrieved from https://app1.fldoe.org/ Comparing the effect of EDPA and FDPA California high school exit examination FCATDemographics/ on university students’ examination results. (CAHSEE). Retrieved from http://www.cde. Florida Department of Education. (2015a). Flori- Malaysian Online Journal of Educational ca.gov/ta/tg/hs/index.asp da’s single-year dropout rates. Retrieved from Technology, 3(1), 28-34. California Department of Education. (2015b). http://www.fldoe.org/core/fileparse.php/7584/ Krashen, S. (2012a, July 25). How Common Core California high school exit exam (CAHSEE) urlt/Dropout-Rate-Report-1314-jtk.pdf will change testing in schools. The Washing- state demographic summary report. Retrieved Florida Department of Education. (2015b). ton Post. Retrieved from https://www.wash- from http://cahsee.cde.ca.gov/ExitProf1.asp?- Graduation requirements for Florida’s ingtonpost.com/blogs/answer-sheet/post/ cLevel=State&cYear=2014-15&cChoice=Ex- statewide assessments. Retrieved from http:// how-common-core-will-change-testing-in- itProf1&cAdmin=C&tDate=000000&Test- www.fldoe.org/core/fileparse.php/3/urlt/ schools/2012/07/25/gJQAzIJB9W_blog.html Type=E&cGrade=10&Pageno=1 fcatpass.pdf Krashen, S. (2012b, July 21). Letter to the editor. California Department of Education. (2015). Florida Department of Education. (2016a). New York Times Sunday dialogue: Improv- Facts about English learners in California. Student enrollment. Retrieved from https:// ing our schools. Retrieved from http://www. Retrieved from http://www.cde.ca.gov/ds/sd/ edstats.fldoe.org/SASWebReportStudio/got- nytimes.com/2012/07/22/opinion/sunday/ cb/cefelfacts.asp oReportSection.do?sectionNumber=1 sunday-dialogue-improving-our-schools. Campitelli, G., & Gobet, F. (2011). Deliberate Florida Department of Education. (2016b). html?_r=0 practice: Necessary but not sufficient. Cur- Students performance results: State reading Layton, L. (2015, October 24). Study says rent Directions in Psychological Science, demographic report (interactive report). standardized testing is overwhelming na- 20(5), 280-285. Retrieved from https://app1.fldoe.org/FCAT- tion’s public schools. The Washington Post. Cardelle-Elawar, M. (1992). Effects of metacog- Demographics/Selections.aspx?reportType- Retrieved from https://www.washingtonpost. nitive instruction on low-achievers in math- ID=6&level=State&subj=Reading com/local/education/study-says-standard- ematics problem solving. Paper presented Gaier, E. L. (1962). Students’ perceptions of ized-testing-is-overwhelming-nations-pub- at the Annual Convention of the American factors affecting test performance. Journal lic-schools/2015/10/24/8a22092c-79ae-11e5- Association for Educational Research, San of Educational Research, 55(10), 561-566. a958-d889faf561dc_story.html Francisco, CA. Geldhof, G. J., Gestsdottir, S., & Stefansson, K. Lazarín, M. (2014). Testing overload in Amer- Center for Disease Control. (2008 July). (2015). Selection, optimization, and compen- ica’s schools. The Center for American Data collection methods for program eval- sation: The structure, reliability, and validity Progress. Retrieved from https://www. uation: Focus groups. ETA Evaluation of forced-choice versus Likert-type measures americanprogress.org/issues/education/ Briefs. Retrieved from http://www.cdc.gov/ in a sample of late adolescents. International report/2014/10/16/99073/testing-over- healthyyouth/evaluation/pdf/brief13.pdf Journal of Behavioral Development, 39(2), load-in-americas-schools/ Center on Education Policy. (2009). State high 171-185. Likert, R. (1932). A technique for the measure- school exit exams: Trends in test programs, Giambo, D. (2010). High stakes testing, high ment of attitudes. Archives of Psychology, alternate pathways, and pass rates. Retrieved school graduation, and limited English 22, 140, 55. from www.cep-dc.org/cfcontent_file.cfm?At- proficient students: A case study. American Lizzo, A., Wilson, K., & Simons, R. (2002). Uni- tachment=CEP_HSEE09Report_up011310. Secondary Education Journal, 38(1), 44-56. versity student perceptions of the learning pdf Hong, E., & Peng, Y. (2008). Do Chinese students’ environment and academic outcomes: Impli- Center on Education Policy. (2012). State high perceptions of test value affect test perfor- cations for theory and practice. Studies in school exit exams: A policy in transition. Re- mance? Mediating role of motivational and Higher Education, 27(1), 27-52. trieved from http://www.cep-dc.org/display- metacognitive regulation in test preparation. Luykx, A., Lee, O., Mahotiere, M., Lester, B., Document.cfm?DocumentID=408 Learning and Instruction, 18, 499-512. Hart, J., & Deaktor, R. (2007). Cultural and Coltrane, B. (2003). English language learners Hopkins, M., Thompson, K. D., Linquanti, R., home language influences on children’s and high-stakes tests. An overview of the Hakuta, K., & August, D. (2013). Fully ac- responses to science assessments. Teachers issues. ERIC Digest. Retrieved from http:// counting for English learner performance: A College Record, 109(4), 897-926. www.ericdigests.org/2003-4/high-stakes.html key issue in ESEA reauthorization. Educa- Mahon, E. A. (2006). High-stakes testing and Cotterell, B. (2015 April 14). New law reduces tional Researcher, 42(2), 101-108. English language learners: Questions of testing at Florida schools. Reuters News Ser- Huang, J., Han, T., & Schnapp, K. (2012). Do validity. Bilingual Research Journal, 30(2), vice. Retrieved from http://www.reuters.com/ high-stakes test really address English 479-497. article/us-usa-florida-education-idUSKBN- language learners’ learning needs? A dis- Maurer, T. J., & Andrews, K. D. (2000). Tradi- 0N52F720150414 cussion of issues, concerns, and implica- tional, Likert, and simplified measures of Council of the Great City Schools. (2015). tions. International Journal of Learning & self-efficacy. Educational and Psychological Student testing in American’s great city Development, 2(1). Retrieved from http:// Measurement, 60(6), 965-73. schools: An inventory and preliminary anal- www.macrothink.org/journal/index.php/ijld/ Merrow, J. (2004). Disappearing dropouts. The ysis. Retrieved from www.cgcs.org/cms/lib/ article/view/1472/1188 Jim Leher News Hour. Retrieved from www. DC00001581/.../87/Testing%20Report.pdf Israel, G. D., & Galindo-Gonzalez, S. (2014). merrow.org/podcast/01.mp3 Davoudi, M., & Sadeghi, N. A. (2015). A sys- Using focus group interviews for planning Murphy, S. (2007). Culture and consequences: tematic review of research on questioning or evaluating extension programs. UF/IFAS The canaries in the coal mine. Research in the as a high-level cognitive strategy. English Extension Document AEC387. Retrieved Teaching of English, 42(2), 228-244. Language Teaching, 8(10), 76-90. from http://edis.ifas.ufl.edu/pdffiles/PD/ National Center for Education Statistics. (2011- De Groot, E. V. (2002). Learning through inter- PD03600.pdf 12). Table 5.5: State high school exit exams, by SPRING/SUMMER 2017 33 Research exam characteristics and state, 2011-12. U.S. 552-580. ready or not? Education Policy Analysis Department of Education, Institute of Edu- Robinson-Cimpian, J. P., Thompson, K. D., & Archives, 16(1). Retrieved from http://epaa. cation Sciences. Retrieved from http://nces. Umansky, I. L. (2016). Research and policy asu.edu/epaa/v16n1/ ed.gov/programs/statereform/tab5_5.asp considerations for English learner equity. Veenman, M.V.J. & Spaans, M.A. (2005). Rela- National Center for Education Statistics. Policy Insights from the Behavioral and tion between intellectual and metacognitive (2015a). Table 204.20: Number and percent- Brain Sciences, 3(1), 129-137. skills: Age and task differences. Learning & age of public school students participating in Sanchez, M. T., Ehrlich, S., Midouhas, E., & Individual Differences, 15(2), 159-176. programs for English language learners, by O’Dwyer, L. (2009). Analyzing performance Veenman, M. V. J., Wilhelm, P., & Beishuizen, J. state: Selected years, 2002-03 through 2012- by grade 10 Hispanic high school students on J. (2004). The relation between intellectual 2013. Retrieved from https://nces.ed.gov/ the Massachusetts State Assessment: Issues and metacognitive skills from a developmen- programs/digest/d14/tables/dt14_204.20.asp and answers. REL 2009-No. 071. Retrieved tal perspective. Learning and Instruction, National Center for Education Statistics. at http://ies.ed.gov/ncee/edlabs/regions/ 14(1), 89-109. (2015b). Table 1. Public high school 4-year northeast/pdf/REL_2009071.pdf Vrieling, E., Bastiaens, T., & Stijnen, S. (2012). adjusted cohort graduation rate (ACGR), by Scholes, R., & Lain, M. M. (1997, March 24-28). Consequences in increased self-regulated race/ethnicity and selected demographics The effects of test preparation activities on learning opportunities on student teachers’ for the United States, the 50 states, and the ACT assessment scores. Paper presented at motivation and use of metacognitive skills. District of Columbia: School year 2013–2014. the Annual Meeting of the American Edu- Australian Journal of Teacher Education, Retrieved from https://nces.ed.gov/ccd/tables/ cational Research Association, Chicago, IL. 37(8). Retrieved from http://files.eric.ed.gov/ ACGR_RE_and_characteristics_2013-14.asp ERIC # ED409341. fulltext/EJ995229.pdf National Center for Education Statistics. (2016). Shammas, B. (2016, March 4). Frustrated south Wallace, K. (2015, April 24). Parents all over The nation’s report card (Interactive report). Florida families opting out of state exams. U.S. ‘opting out’ of standardized student Retrieved from http://www.nationsreport- Sun Sentinel. Retrieved from http://www. testing. CNN. Retrieved from http://www.cnn. card.gov/dashboards/report_card.aspx# sun-sentinel.com/local/palm-beach/fl-opt-out- com/2015/04/17/living/parents-movement- Page-Bucci, H. (2003). The value of Likert scales movement-20160304-story.html opt-out-of-testing-feat/ in measuring attitudes of on-line learners. Solórzano, R. W. (2008). High stakes testing: Is- Welsh, M. E., Eastwood, M., & D’Agostino, J. Retrieved from http://www.hkadesigns.co.uk/ sues, implications, and remedies for English V. (2014). Conceptualizing teaching to the websites/msc/reme/likert.htm language learners. Review of Educational test under standards-based reform. Applied Patrick, H. & Middleton, M. J. (2002). Turning Research, 78(2), 20-329. Measurement in Education, 27(2), 98-114. the kaleidoscope: What we see when self-reg- Texas Education Agency. (2014). Enrollment in WESH Orlando. (2015, April 14). Florida gov- ulated learning is view with a qualitative Texas public schools 2013-2014. Retrieved ernor signs bill to limit standardized tests. lens. Educational Psychologist, 37(1), 27-40. from http://tea.texas.gov/acctres/enroll_in- Retrieved from http://www.wesh.com/news/ Purpura, J. E. (1997). An analysis of the rela- dex.html florida-gov-signs-bill-to-limit-standardized- tionships between test-takers’ cognitive and Texas Education Agency. (2015). Fall 2015 EOC: tests/32360186 metacognitive strategy use and second lan- Combined summary reports. Retrieved from Wismath, S., Orr, D., & Good, B. (2014). Meta- guage test performance. Language Learning, http://tea.texas.gov/Student_Testing_and_ cognition: Student reflections on problem 47(2), 289-325. Accountability/Testing/State_of_Texas_As- solving. Journal of Excellence in College Ravich, D. (2011). The death and life of the sessments_of_Academic_Readiness_(STA- Teaching, 25(2), 69-90. great American school system: How testing AR)/STAAR_Statewide_Summary_Re- Zhu, W., & Flaitz, J. (2005). Using focus group and choice are undermining education. New ports_2015-2016/ methodology to understand international York: Basic Books. Texas Education Agency. (2016a). State assess- students’ academic language needs: A com- Ravitch, D. (2016a, January 29). District ad- ments: Past, present, and future. Retrieved from parison of perspectives. Teaching English ministration poll: Significant resistance tea.texas.gov/WorkArea/linkit.aspx?Link- as a Second or Foreign Language, 8(4). Re- to testing. Retrieved from http://dianerav- Identifier=id&ItemID=25769825328&li- trieved from http://tesl-ej.org/ej32/a3.html. itch.net/2016/01/29/district-administra- bID=25769825424 tion-poll-significant-resistance-to-testing/ Texas Education Agency. (2016b). State gradua- Ravitch, D. (2016b, January 26). Fairtest: tion requirements. Retrieved from http://tea. The fight against misuse of testing contin- texas.gov/graduation.aspx ues. Retrieved from http://dianeravitch. Tok, S. (2013). Effects of the know-want-learn net/2016/01/26/fairtest-the-fight-against- strategy on students’ mathematics achieve- misuse-of-testing-continues/ ment, anxiety and metacognitive skills. Roberts, F. W., & Dansereau, D. F. (2008). Study- Metacognition and Learning, 8(2), 193-212. ing strategy effects on memory, attitudes, Tsang, S. L., Katz, A., & Stack, J. (2008). Achiev- and intentions. Reading Psychology, 29(6), ing testing for English language learners, MULTICULTURAL EDUCATION 34