|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 PREDICTING HIGH SCHOOL GRADUATION FOR LATINO MALES USING EXPECTANCY VALUE THEORY OF MOTIVATION AND TENTH GRADE READING ACHIEVEMENT SCORES By ERIN OAKLEY KNAPE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010
2 2010 Erin Oakley Knape
3 To my family
4 ACKNOWLEDGMENTS This dissertation could not have been completed without the support, guidance, encouragement, and feedback from many family, friends, colleagues, and professors. First, I would like to thank my doctoral committee chair, Dr. Mary Ann Clark, for her constan t encouragement, gentle guidance, and thoughtful feedback throughout the dissertation process. My departmental committee members, Dr. Ellen Amatea and Dr. Cirecie West Olatunji, have been instrumental in my development over the past several years as a counselor, as an educator, and as a researcher. Additionally, I would like to thank my external committee member, Dr. James Algina, for his patient instruction and guidance through the complex statistical world of MPLUS and structural equation modeling. I w ill be forever grateful for the expertise these professors have shared and the professional example they have set. I would also like to thank Candy Spires and Patty Bruner for the invaluable assistance they have provided me throughout my years in the department. I would also like to recognize the support of my friends and colleagues, without whom this process would have been much more difficult and lonely. While many people (too many to mention) have encouraged me along the way, I would like to recogniz e two in particular. Kelly McWhorter has always been able to raise my spirits and make me laugh, even during the most stressful times of my dissertation, and I feel extremely fortunate for her friendship. I am also so grateful to Dr. Blaire Cholewa for h er honest and genuine friendship over the years. I know I can share my successes and failures with her and can depend on her for an understanding ear, a reality check, or an excited cheer, whichever is needed at the time. I am truly blessed to have such amazing and
5 supportive friends to always remind me that I can accomplish my goals, and to provide much needed diversions along the way. Finally, I would like to recognize the love and encouragement I have received from my family. Even though they haven t always understood what exactly my field of study is or what my research is about, they have always stood behind me and challenged me to grow personally and professionally. My parents have given me many opportunities in my life that have allowed me to be where I am today, and I am extremely blessed and grateful for their love and guidance. I am so fortunate for the relationships I have with my parents and my brother, who have together formed the strong foundation from which I have been able to grow and t hrive. Last, and absolutely not least, I would like to thank my husband, Rob, for his unending love and encouragement. Words cannot express how grateful I am to have had the support of my best friend and partner through the ups and downs of this process. Finally, I would like to recognize the influence of my soonto beborn son, who has already given me so much joy and hope, and who has provided the last little bit of motivation I needed to finally finish this work. It is only through the support, encouragement, and love of this amazing support system that I have been able to complete this dissertation. Above all, I thank God for the many blessings and gifts He has given me. I can only hope to be able to bless others in my life as much as I have been blessed.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 TABLE OF CONTENTS .................................................................................................. 6 LIST OF TABLES ............................................................................................................ 9 LIST OF FIGURES ........................................................................................................ 10 DEFINITION OF TERMS .............................................................................................. 11 ABSTRACT ................................................................................................................... 13 CHAPTER 1 INTRODUCTION .................................................................................................... 15 Statement of the Problem ....................................................................................... 16 Achievement of Males ...................................................................................... 17 Race, Gender, and Poverty .............................................................................. 19 Student Motivation ............................................................................................ 20 High School Requirements ............................................................................... 22 Reading Achievement ...................................................................................... 23 Educational Attainment .................................................................................... 24 Significance of the Study ........................................................................................ 25 School Counselors Role .................................................................................. 25 Potential Implications of This Study .................................................................. 26 Purpose of the Study .............................................................................................. 27 Research Questions ............................................................................................... 28 Hypotheses ............................................................................................................. 28 Theoretical Framework ........................................................................................... 28 Methodology for the Study ...................................................................................... 29 2 LITERATURE REVIEW .......................................................................................... 31 Problem of Achievement Gap ................................................................................. 31 Gender Achievement Gap ................................................................................ 32 Cultural Achievement Gap ................................................................................ 35 Dropout and retention rates ....................................................................... 36 Post secondary enrollment and attainment ................................................. 37 Socioeconomic Achievement Gap .................................................................... 38 Masculinity and Intersectionality ............................................................................. 43 Previous Approaches to Addressing Achievement Gaps ........................................ 45 Expectancy Value Theoretical Framework ............................................................. 46
7 Expectancies .............................................................................................. 47 Task values ................................................................................................ 52 3 METHODOLOGY ................................................................................................... 55 Education Longitudinal Study of 2002 .................................................................... 55 Participants ............................................................................................................. 58 Access .................................................................................................................... 59 Research Variables ................................................................................................ 60 Dependent Variable .......................................................................................... 60 Independent Variables ..................................................................................... 60 Expectancy variables ................................................................................. 62 Task value variables .................................................................................. 63 Reading achi evement ................................................................................ 64 Data Analysis .......................................................................................................... 65 4 RESULTS ............................................................................................................... 68 Descriptive Statistics ............................................................................................... 68 Factor Analysis Results .......................................................................................... 72 Structural Regression Analysis ............................................................................... 73 First Research Question ................................................................................... 76 Signific ant predictive relationships ............................................................. 79 Additional notable relationships and findings ............................................. 80 Second Research Question .............................................................................. 82 5 DISCUSSION ......................................................................................................... 86 Discussion of Results .............................................................................................. 86 Predicting High School Graduation .................................................................. 87 Impact of Reading Achievement ....................................................................... 88 Academic Attitudes and Attainment .................................................................. 90 Expectancy Value Variables ............................................................................. 90 Challenges and Barriers ......................................................................................... 91 Poverty ............................................................................................................. 92 English Proficiency ........................................................................................... 93 Other Barriers ................................................................................................... 94 Additional Challen ges for Latino Males ............................................................ 96 Discrimination ............................................................................................ 97 Need to support family ............................................................................... 97 Implications for School Counselors ......................................................................... 98 School Counselors as Educational Leaders ..................................................... 99 School Family Partnerships ............................................................................ 102 Limitations of the Study ......................................................................................... 105 Directions for Future Research ............................................................................. 106 Conclusion ............................................................................................................ 109
8 REFERENCES ............................................................................................................ 111 BIOGRAPHICAL SKETCH .......................................................................................... 125
9 LIST OF TABLES Table page 2 1 Perce ntage distribution of students across reading achievement levels, by race/ethnicity and grade: 2007 ........................................................................... 36 3 1 Percentage of high school sophomores, by students highest level of education expected: 2002 .................................................................................. 59 3 2 Variable label, name, and description ................................................................. 61 4 1 Descriptive Statistics for Variables in Model ....................................................... 69 4 2 Correlation matrix for variables included in the model ........................................ 70 4 3 Predictive associations retained in the reduced model ....................................... 77 4 4 Statistically significant correlations in final model ............................................... 81 4 5 Comparison of twotailed P values for lower and higher SES 10th grade Latino males ....................................................................................................... 83
10 LIST OF FIGURES Figure page 4 1 Hypothesized model based on Expectancy Value Theory .................................. 74 4 2 MPLUS model of predictive associations among selected ELS:2002 variables for 10th grade Latino males ................................................................................. 78
11 DEFINITION OF TERMS Achievement gap The difference in academic success between two or more groups of students, often identified by gender, race, or socioeconomic status. Culturally Diverse While this term can refer to a population in which a broad range of cultural and ethnic backgrounds ar e represented, in this study this term refers to individuals who identify as Black/African American or Hispanic/Latino American and who are most often impacted by academic achievement gaps (Cholewa & West Olatunji, 2008). Culture/ethnicity Based on the fe deral standards for identifying ones racial or cultural background. The national database used in this study, the ELS:2002, uses the categories of White, Black/African American, Asian or Pacific Islander, including Native Hawaiian, American Indian or Alaska Native, Hispanic or Latino, or Multiracial (Ingels & Scott, 2004). This database allows students to select more than one cultural group. However, professional counseling literature often describes ethnic categories differently, using the cultural descriptors of African American and Latino in place of the federal racial categories of Black and Hispanic (e.g. Darling Hammond, 2000; Ingels & Scott, 2004; Llagas & Snyder, 2003). In this study, the terms Black and Hispanic will be used interchangeably with African American and Latino, respectively, to reflect the sources from which information has come. Expect ancy how well a person anticipates performing on a given task or activity (Eccles et al., 1983). Gender often considered on a spectru m including masculine and feminine, in this study this term is used synonymously with sex to refer to being male or female. Intersectionality the combination or confluence of multiple social identities (i.e. ethnicity, socioeconomic status, sexual orient ation, social roles) for an individual (Hurtado & Sinha, 2008). Low income a demographic indicator that shows if a family is situated in the bottom 20% of all family incomes in the United States (Planty et al., 2008). The Department of Education also de fines low income schools as those where over 75% of students participate in the free or reduced lunch program (Planty et al., 2008). The National School Lunch Program of the Department of Agricultures Food and Nutrition Services department provides free lunches to students whose families have incomes less than 130% of the poverty level, and reducedprice lunches to students whose families earn less than 185% of the poverty level (U.S. Department of Agriculture,
12 2009). For the 20092010 school year, these income levels for a household of 4 people were approximately $28,665 to receive free lunches and $40,793 to receive reduced price lunches (U.S. Department of Agriculture, 2009). Often, researchers will use free and reduced lunch status to indicate povert y or SES level (e.g. Taylor & Graham, 2007). Masculine/Masculinity the qualities, character, or essence of being male, usually described in contrast with feminine qualities or femininity; a socially constructed set of values, practices, and roles for boys and men in a given society (Connell, 1995). Motivation a persons desire and drive to accomplish a goal, such as succeeding in school or making good grades, and which is impacted by a variety of individual and contextual experiences (see Eccles et al, 1983). Sex classified as male or female, this variable is based on genetic and physiological characteristics, and is primarily biological in nature. The national database used in this study asks respondents to select whether they are male or female. In this study, sex is used synonymously with gender. Socioeconomic status (SES) comprised of a persons economic, social, cultural, and physical environments, typically stemming from a combination of family income, education level, demographics, and social positi oning within a culture. The national database used in this study includes family income, parental education levels, and parental occupation in the composite SES variable, which is divided into quartiles (BYSES2QU). Task value the importance a person places on achieving a certain goal or succeeding at a given task, which is impacted by ones cultural and social contexts (Eccles et al., 1983). Underachievement the phenomenon that occurs when a student possesses the intellectual potential to succeed academically but the students grades or school performance fall short of these expectations.
13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doct or of Philosophy PREDICTING HIGH SCHOOL GRADUATION FOR LATINO MALES USING EXPECTANCY VALUE THEORY OF MOTIVATION AND TENTH GRADE READING ACHIEVEMENT SCORES By Erin Oakley Knape December 2010 Chair: Mary Ann Clark Major: Mental Health Counseling National education data indicate that young men of color and students living in poverty are not experiencing the same academic success as their female, White, or higher socioeconomic status peers, as evidenced by low reading achievement levels and high dropout rates. Of particular concern is the underachievement of Latino males, who currently have the lowest high school completion and postsecondary enrollment rates of any student population in the country. Latino students represent the largest and fastest growing demographic of culturally diverse school age children, and Latino males in particular face many institutional challenges and barriers that negatively impact their motivation and achievement. In their unique position of leadership and with a focus on the contexts impacting students lives, school counselors represent key participants and advocates in the understanding and improvement of the educational experiences of Latino young men. The purpose of this study is to use Eccles et al. (1983) Expec tancy Value Theory of motivation and tenth grade reading achievement scores to identify the factors that predict high school graduation for Latino males from various socioeconomic levels.
14 Data from 1099 10th grade Latino male respondents from the national ly representative Education Longitudinal Study of 2002 were used, and structural regression analyses were performed using MPLUS computer software. Results indicate that high school graduation was significantly predicted by students perceptions that parents expect success in school and by students English teachers educational expec tations for academic attainment. Additionally, student reading achievement scores significantly predicted student academic expectations for lower, but not higher, socioeconomi c Latino males. Implications for school counselors work with Latino males are discussed, and directions for continued research are proposed.
15 CHAPT ER 1 INTRODUCTION Educators and policymakers have given much attention to decreasing and eliminating gender, cultural, and socioeconomic gaps in educational achievement and attainment in the United States. Specifically, national data indicates that male students of color and students in poverty are not experiencing the same academic success as their female, White, or hig her income peers (Bemak & Chung, 2005; Education Trust, 2005a, 2005b; KewalRamani, Gilbertson, Fox & Provasnik, 2007; Planty et al., 2008). Latino students represent the largest and fastest growing ethnic group in this country, and Hispanic males have the highest dropout rates of any student population in the U.S. (Bernstein, 2008; Planty et al., 2008; U.S. Census Bureau, 2008). Latino students are also three times more likely than White students to live in poverty and attend high poverty schools (Prelow & Loukas, 2003). To tackle these existing gender, cultural, and socioeconomic achievement gaps, the No Child Left Behind Act (NCLB) mandated that all educators, including school counselors, focus their efforts on the academic success of all students, with particular emphasis on multicultural and low income children (U.S. Department of Education, 2001). School counselors play a key role in the educational process and in ensuring that all students academic needs are being met by the school (American School Counseling Association, 2005; Bemak, 2000; Bryan, 2005; Stone & Dahir, 2006). In addition, school counselors are in a unique leadership position to facilitate the connection between schools, families, and communities (Amatea & West Olatunji, 2007a; Clark & Stone, 2007; Stone & Dahir, 2006; Walsh, Barrett, & DePaul, 2007) with the understanding of the multiple influences on student achievement and attainment,
16 based on Bronfenbrenners ecological systems model (1986). There has also been a shift away from a deficit model of understanding student achievement to an emphasis on resilience and strengths based counseling that stresses both the importance of partnerships among schools, families, and communities and the appreciation of student cultural and social resources (Amatea, SmithAdcock, & Villares, 2006; Bryan, 2005; Galassi & Akos, 2007; Walsh et al., 2007). It is this unique position of leadership and a focus on the contexts surrounding students that allows school counselors to represent key participant s and advocates in the understanding and improvement of the educational experiences of Latino young men. Statement of the Problem In recent years, the topic of male underachievement has received attention from popular media and researchers alike, increas ing the publics awareness of this problem in our countrys schools. More often, parents hear their sons teachers say that their child is not achieving to his full potential or that he just doesnt seem motivated (Gurian & Stevens, 2005; Lopez, 2003; Sax, 2007). National studies lend support to teacher comments like these, showing that boys are falling far behind their female counterparts in several indicators of academic success including grades, test scores, discipline referrals, special education referrals, and dropout and graduation rates (Clark, Flower, Walton, & Oakley, 2008; Clark, Oakley, & Adams, 2006; Education Trust, 2005a; National Center for Education Statistics [NCES], 2006; The College Board, 2010). The largest gap in achievement between genders exists in the areas of reading and writing, as evidenced in educational literature (Bailey & Paisley, 2004; Blackhurst & Auger, 2008; Blair, Blair, & Madamba, 1999; Freeman, 2004; Gorard & Smith, 2004; Moss, 2000) and through the National Assess ment of Educational Progress (NAEP),
17 national standardized exams issued by the U.S. Department of Education (NCES, 2005, 2006, 2007; Planty et al., 2008). Furthermore, significant gender gaps in educational attainment, namely high school graduation, are c hallenging for school counselors and other education stakeholders (NCES, 2006, Planty et al., 2008). Clearly, schools are not meeting the needs of all of their students equally. Achievement of Males When considering achievement data, it is important for school counselors and other educators to understand the underachievement of males in the context of other gaps in educational success. It is essential to understand which students, and which males, are being underserved by the schools to best address the achievement gaps (Cholewa & West Olatunji, 2008). The term achievement gap is used loosely to describe differences in academic achievement between groups of people, historically among students of various ethnicities, socioeconomic status (SES), and gen der. The gender and cultural achievement gap in reading and writing in primary and secondary schools in the United States has been documented for many years, as well as the gap between lower and higher socioeconomic status (Childrens Defense Fund [CDF], 2005; Freeman, 2004; Gorard & Smith, 2004; NCES, 2007; Planty et al., 2008). Similar gender, cultural, and socioeconomic gaps in educational attainment have also been well documented through high school and postsecondary graduation rates with males, students of color, and lower income students graduating high school at significantly lower rates than their respective female, White, and upper class peers (Amatea & West Olatunji, 2007b; Education Trust, 2005a; KewalRamani et al., 2007; Planty et al., 2008; T he College Board, 2010). Underscoring the prevalence of these achievement and attainment gaps, a survey of urban school leaders found that 80% of
18 these leaders ranked closing achievement gaps as the primary concern facing their school districts (Snipes, W illiams, & Petteruti, 2006). Professional school counseling and education literature has drawn significant attention to the Black White achievement gaps, with relatively less attention being given to the educational experiences of Latino students (DotsonBlake, Foster, & Gressard, 2009; Saenz & Ponjuan, 2009). Hispanic males achieve far below their female counterparts on national standardized exams, and drop out of school at a significantly higher rate than White males (NCES, 2007). For example, Latinos have the highest dropout rates of any population, with approximately 22% not completing high school, in comparison to 6% of White students and 11% of African American students (Planty et al., 2008). The underachievement of Latino students is concerning for school counselors and other educators, particularly given the shifting demographics of the American school population. The United States is becoming more ethnically diverse each year, and census data projects that by the year 2023, more than half of all children will be nonWhite (U.S. Census Bureau, 2008). Latinos represent the largest and fastest growing of these non White student populations, with almost 34% of the Hispanic population being under the age of 18 (Bernstein, 2008; U.S. Census Bureau, 2008). This increasing diversity is also reflected in the general population (adults and children), with a total of 45.5 million Hispanic individuals making up 15% of the total U.S. population as of 2007 (Bernstein, 2008). Furthermore, it is expected that non White individuals will comprise 54% of the country by the year 2050, thus representing a majority of the population (U.S. Census Bureau, 2008).
19 Race, Gender, and Poverty In addition to the documented cultural achievement gaps, boys from lower socioeco nomic groups perform less well than girls from lower SES groups on exams nationally and internationally (Francis & Skelton, 2005; NCES, 2007; OECD, 2003). In fact, the gap between boys and girls in reading is even wider for economically disadvantaged students than for middle class students (Francis & Skelton, 2005). Students of color, including Latino students, are overrepresented in lower socioeconomic groups (Prelow & Loukas, 2003). In addition to the increasing cultural diversity of the nations popul ation, Planty and his colleagues (2008) report in the U. S. Department of Educations annual Conditions of Education publication that the number of elementary and secondary students living below the poverty line has increased since 1979, with approximately 17% of students nationwide in poverty in 2006 (see also Berliner, 2006). Of special concern is the rapidly growing Latino student population, who are three times as likely as White students to live in poverty (Prelow & Loukas, 2003), with an estimated 28% of Latino children living below the poverty line (Llagas & Snyder, 2003). Often, student poverty and low SES status is indicated in school systems as eligibility for and participation in the free or reduced lunch program, which provides meals to student s whose families are situated below or just above the poverty level (Berliner, 2006; Planty et al., 2008; U.S. Department of Agriculture, 2009). Clearly, student poverty is prevalent in the United States in general, and in the Latino population specifically. While certainly many Latino males and many males from lower SES backgrounds are doing very well in school, school counselors need to be aware of the magnitude of the current and persisting achievement gaps concerning Latino students to effectively pa rtner with other stakeholders (teachers, administrators,
20 parents, etc.) to improve the education of this sizeable and underserved group of students (Amatea & West Olatunji, 2007b; Berliner, 2006). Student Motivation Motivation in school can impact short term goals such as grade completion and high school graduation (Dimmitt, 2003). Moreover, decreased motivation can potentially dampen other long term educational and life goals, such as attending college or finding a fulfilling career (Kobrin, Patterson, Shaw, Mattern, & Barbuti, 2008; McCall, 1994; The College Board, 2008). Attitudinal and motivational data suggest that, as a group, boys do not think school or reading are as important to their lives as do girls (Clark et al., 2006; Flowers, 2003; S chwartz, 2002). For example, a national longitudinal study by NCES revealed that girls reported their coursework was more meaningful, interesting, and more applicable to their futures when compared to their male peers (NCES, 2005). When students do not s ee the relevance of schoolwork to their future, they lack the motivation to put forth the effort necessary to succeed academically (Carey & Martin, 2007; Dimmitt, 2003), and lower achievement motivation has been associated with underachieving students (Preckel, Holling, & Vock, 2006). However, school counselors have come to view motivation (and other related constructs) as not simply a standalone intrapsychic characteristic, but rather as being influenced by the various contexts and systems that surround a student (Amatea & Sherrard, 1994; Neill & Kniskern, 1982). Furthermore, Gordon Rouse & Austin (2002) report that motivation is demonstrated differently across ethnic groups and genders, perhaps in part because of students differential experiences in the school context. Several researchers have also pointed to a cultural mismatch, or discontinuity, between students of color, low SES, or male students and the school settings they attend as a significant and essential
21 consideration when discussing the academic performance and motivation of these students (Cholewa & West Olatunji, 2008; Howard, 2003). Additionally, Latino students race gender experiences accumulate and ultimately affect how men and women come to understand the role of education in their lives as well as their prospects for social mobility (Lopez, 2003, p. 6), pointing again to students perceived value of (and motivation for) academic attainment. When schools repeatedly devalue certain cultural values and experiences within the classroom setting, Latino students, and males in particular, often disengage from the schooling process over time, and may become less motivated to succeed in that setting (Lopez, 2003). Clearly, motivation is shaped by ongoing race and gender processes in the cl assroom, in addition to other systemic experiences. School counselors need to gain a greater understanding of these gender, cultural, and socioeconomic differences in the expression of motivation toward school success in order to best serve these young men. Researchers and school counselors must also consider how Latino males social identities and attitudes influence their achievement motivations and behaviors. Previous studies have focused on the prevalence of and the negative effects of boys attitud es that school is not macho or cool (Francis, 2000; Francis & Skelton, 2005; Sax, 2007), or is not socially acceptable for them (Van Houtte, 2004). Boys social culture in school is extremely salient in the minds of male students, and this social cult ure can vary across ethnic groups as well as socioeconomic levels (Howard, 2003; Lopez, 2003). Researchers have indicated that the need to be accepted by peers can influence or even outweigh young mens desire to succeed academically (Van De Gaer, Pustjens, Van Damme, & De Munter, 2007).
22 High School Requirements High school represents a key time in the academic, social, and attitudinal development of adolescents. States conduct mandated achievement tests throughout the K 12 schooling process, but the tests during high school bear particular consequences on a students academic progress (Center on Education Policy, 2008). For example, many states require that students demonstrate proficiency on the 10th grade achievement tests in order to graduate from high school (Florida Department of Education, 2007). In fact, by the year 2012, 26 states will require high school graduates to demonstrate proficiency on their statemandated high school exit exams to graduate, and these exams are generally based on 9th o r 10th grade competencies (Center on Education Policy, 2008). Furthermore, these 26 states will comprise 74% of the countrys public high school students, indicating the significant impact that these exit examinations will have on American education (Cent er on Education Policy, 2008). Considering these relatively significant consequences for the 10th grade examinations in particular, this age group represents an ideal time for examining the impact of reading achievement and motivation on graduation. In addition to the testing consequences of 10th grade, mandatory schooling ends at age 16, and thus represents a time when students may drop out of school. Data from the ELS:2002 indicate that approximately 5% of students do not complete high school because t hey drop out, and that 65% of the 10th grade dropouts were male for the 20022003 school year (HampdenThompson, Warkentien, & Daniel, 2009). Students motivation and achievement at the 10th grade level can have serious implications for future educational attainment.
23 Reading Achievement While achievement in all academic areas is encouraged and is important to the success of a student in high school, reading achievement appears to have significant implications for success in school and beyond (Kobrin et al., 2008; The College Board, 2008). As previously discussed, reading and writing represent the areas where, as a group, males tend to fall significantly behind their female counterparts at all grade levels, both in the United States and internationally (F rancis & Skelton, 2005; Freeman, 2004; Gorard & Smith, 2004; Moss, 2000; Planty et al., 2008). Students use their literacy skills beyond just their English class; being able to read and write well is vital to student success in other subjects as well, as many primary, secondary, and post secondary educational settings have adopted writing (or language) across the curriculum initiatives. As evidence of the desirability of this key skill, an additional writing component was added in 2005 to the College Boar ds Scholastic Aptitude Test (SAT), one of the two major college entrance examinations in the United States. The addition of this writing component, supplementing the mathematics and verbal sections, emphasizes the nations focus on the importance of literacy skills beyond the high school years into college or the working world (Kobrin et al., 2008). Furthermore, validity studies indicate that students writing scores on the SAT were the best predictor of first year college grade point averages when compared to the other two sections of the SAT (Kobrin et al., 2008). Clearly, reading and writing are fundamental skills, not only in the completion of high school, but also in success in college and beyond. Therefore, the achievement gaps in reading and writ ing for Latino males across socioeconomic levels are of particular concern for high school graduation as well as for the long term success of these students.
24 Educational Attainment Given the general academic, attitudinal, and behavioral differences descr ibed by the literature thus far, school counselors need to consider the particular impact of Latino underachievement on educational attainment, post secondary education, and entrance into the workforce for these young men. First, underachievement by culturally diverse males and low SES males in high school has implications for postsecondary attainment (Kelly, Schneider, & Carey, 2010; Planty et al., 2008). Echoing an ecosystemic perspective, Lopez suggests that the impact of the high school context on Lat ino students can have lasting effects on the value of education in general, stating that ordinary day to day school practices and classroom dynamics are racialized and gendered and in turn shape youth views about the role of education in their lives (20 03, p. 68). Women currently represent the majority of all college students and earn a majority of the academic degrees offered, and the percentage of women attending college is increasing relative to men (Adebayo, 2008; Peter, Horn, & Carroll, 2005). Men of color are significantly underrepresented in higher education enrollments, and males earn only about a third of postsecondary degrees awarded to Hispanic students, a pattern similar to that of African American and Native American students (Blackhurst & Auger, 2008; Saenz & Ponjuan, 2009). The extremely low postsecondary completion rates of Hispanic males relative to Hispanic females and to males from other ethnic groups is cause for national attention (Kelly et al., 2010). The concern surrounding underachievement also reaches beyond the school years into adulthood. McCall (1994), who studied high school underachievers into their adult lives, found that over a decade after high school graduation, individuals educational att ainment and job status was more consistent with their school grades than
25 their actual abilities meaning that they never really caught up to their potential Additionally, McCall found that high school underachievers generally demonstrated less persist ence in completing their college degrees, holding on to their jobs, and maintaining their marriages (1994, p. 18). Evidently, the disproportionate rates of educational attainment for underachieving Latino males warrant additional investigation to ensure this growing population attains a more positive future (Saenz & Ponjuan, 2009). National data and professional research clearly indicates the growing concern of Latino male underachievement through academics, attitudes, and educational attainment (Francis & Skelton, 2005; Freeman, 2004; Jones & Myhill, 2004; NCES 2007; Saenz & Ponjuan, 2009; Van Houtte, 2004). Significance of the Study The Education Longitudinal Study of 2002, a national education database, will be investigated to isolate the motivation factors and reading achievement scores that predict high school graduation for Latino males across socioeconomic levels. By identifying these predictive factors within various socioeconomic groups, the findings may contribute to school counselors and oth er educators understanding of the significant gender achievement gaps in graduation for Latino students. The No Child Left Behind Act acknowledges the existing achievement gaps and mandates that all educators, including school counselors, strive for the academic success for all students, with particular attention to culturally diverse and low income learners (U.S. Department of Education, 2001). School Counselors Role All educators, and particularly school counselors, need to be aware of the variety of contextual and systemic factors which impact the motivation and achievement of the
26 students within their schools. Based on the American School Counselor Association (ASCA) National Model for school counseling programs (2005), school counselors are encour aged to promote success for all students by integrating social, psychological, career, and academic development. Additionally, ASCA (2005) asserts that school counselors are in a unique position to call attention to situations where the school system is n ot meeting the academic needs of a given student population, such as Latino males. For example, by examining data that demonstrate achievement gaps between Latino and White students, the school counselor in a leadership position can seek to understand those students needs and how the school is failing to meet those needs, and can collaborate with other stakeholders to make necessary changes (Amatea & West Olatunji, 2007a; Bemak, 2000; Bryan, 2005; Clark & Stone, 2007; Lee, 2001; Smith Adcock, Daniels, Lee, Villalba, & Indelicato, 2006; Stone & Dahir, 2006). Additionally, one commonly discussed focus of school counselor teacher collaboration, where counselors can bring their skills in multicultural competency and advocacy, is to develop culturally respons ive and relevant pedagogy within the classrooms (Bemak & Chung, 2005; Cholewa & West Olatunji, 2008; Goh, Wahl, McDonald, Bissett, & Yoon, 2007; Lee, 2001). Stanard (2003) also asserts that school as well as community counselors need to be concerned with and actively involved in the efforts to establish culturally responsive school policies and programs, as well as to develop connections among schools, communities, and families to promote academic success for students of color in particular. Potential Im plications of This Study With an increased focus on closing achievement gaps between genders, ethnicities, and varying socioeconomic statuses of students, there is a need for school
27 counselors, other educators, and researchers to further understand withingroup as well as betweengroup differences (Gordon Rouse & Austin, 2002; Meece et al., 2006). Despite the increasing cultural diversity of the school aged population, little research has examined how gender differences in motivation differ by ethnicity, race, or socioeconomic status (Meece et al., 2006, p. 365; Meece & Kurtz Costes, 2001). Similarly, limited research has been conducted focusing on Latino achievement motivations across various socioeconomic levels. The results from this study have potential implications for educational policy coinciding with the national push to incorporate evidencebased and culturally responsive practices into schools across the country. The nationally representative Education Longitudinal Study of 2002 will allow the results of this research to generalize to large segments of the population, making the findings more relevant and robust than if smaller, more regional samples were used. Additionally, this study may provide support for a theoretical lens through which s chool counselors may view and understand the issue of Latino male underachievement across socioeconomic groups. Finally, school counselors, teachers, and administrators may use the results of this study to inform practice through the interventions, teachi ng strategies, and school policies they employ aimed at raising all students achievement and attainment. Purpose of the Study The purpose of this study is to use Eccles et al. (1983) Expectancy Value Theory of motivation and tenth grade reading achievem ent scores to identify the factors that predict high school graduation for Latino males from various socioeconomic levels.
28 Research Questions The following research questions will be addressed in this study: 1. How do motivational factors and tenth grade r eading achievement scores predict high school graduation for Hispanic males? 2. To what extent do motivational factors and tenth grade reading achievement scores differ among Hispanic males from varying levels of SES? Hypotheses The following null hypotheses correspond to the research questions listed above: 1. Student motivational factors and tenth grade reading achievement scores do not predict student high school graduation for Hispanic males. 2. There is no difference in motivational factors and tenth grade reading achievement scores among Hispanic males from varying levels of SES. Theoretical Framework The achievement motivations of high school Latino males will be examined from the framework of Expectancy Value theory of motivation proposed by Eccles and her colleagues (1983), and this theory will guide the selection and operationalization of variables in this study. This theory posits that motivation is influenced by expectancy and task value beliefs, and that these beliefs impact a persons performance or achievement (Eccles et al., 1983; Wigfield & Eccles, 2002). According to Eccles and her colleagu es, expectancies are peoples beliefs about how well they will do on a given task, and are influenced by beliefs such as self concept of ability, perception of task difficulty, perceptions of others expectations, causal attributions, and locus of control (Eccles et al., 1983; Wigfield & Eccles, 2002). Task values, on the other hand, are influenced by the attainment value of the task, the intrinsic or interest value of the task, and the utility value of that task for achieving a future goal (Eccles et al., 1983).
29 This theory suggests that there are multiple factors that make up the broader concept of motivation, and that these factors can be based on personal experiences, perceptions, and ones social and cultural context. The inclusion of cultural and s ocial influences in this theory echoes the emphasis on context and environment that is characteristic of ecosystemic theories (Amatea & Sherrard, 1994; Amatea & West Olatunji, 2007a; Bronfenbrenner, 1986). In this way, motivation can be viewed in this stu dy not as a simple intrapsychic characteristic, but rather as a set of factors that are influenced largely by environment, experience, and context (Amatea & Sherrard, 1994; Neill & Kniskern, 1982). Previously, Expectancy Value theory has been applied freq uently to motivation and achievement in mathematics, but not as often with reading, graduation, or Latino students (Meece, Glienke, & Burg, 2006). Specifically, this study will use the constructs from Expectancy Value Theory in addition to students readi ng achievement scores to understand the achievement motivations and educational attainment of Latino males. This theory will be explored further in Chapter 2 (Literature Review). Methodology for the Study Data from the Education Longitudinal Study of 2002 (ELS:2002) will be analyzed to understand how high school Latino males motivations for academic success and their reading achievement scores predict graduation from high school. The ELS:2002 database will be used for two main reasons. First, the ELS: 2002 survey is a longitudinal study consisting of a large, nationally representative sample of participants. Second, the students surveyed in the ELS:2002 represent more recent high school students than the participants of earlier Department of Education longitudinal studies, and would therefore represent more current trends in educational motivation and achievement.
30 The variable selection and data analysis for this study will be guided by Eccles Expectancy Value Theory (1983). Confirmatory factor analy sis will be used to group questionnaire items into factors representing constructs from Expectancy Value Theory, and these factors, in addition to students reading achievement scores in 10th grade, will be used in a structural logistic regression model to predict high school graduation for Latino male students. The data will also be examined to assess significant differences in predictive motivation factors and reading achievement scores across varying levels of SES for these Hispanic males. Further details about the Education Longitudinal Study of 2002 and the data analysis can be found in the methodology chapter (Chapter 3). The following chapters will provide a more thorough explanation of the background of the current achievement gap for Latino males and the way this problem will be investigated in this study. Chapter 2 will contain a review of the relevant literature and an explanation of the theoretical framework that guides the study. Chapter 3 will explain the research methodology and data anal ysis procedures to be used, as well as provide further information about the Education Longitudinal Study of 2002. Chapter 4 will review the results of the data analyses, and Chapter 5 will contain an indepth discussion of the results and the implications of the findings.
31 CHAPTER 2 LITERATURE REVIEW The purpose of this study is to use Eccles et al. (1983) Expectancy Value Theory of motivation and tenth grade reading achievement scores to identify the factors that predict high school graduation f or Latino males from various socioeconomic levels. This chapter contains a review of the evidence and literature relevant to the present investigation predicting graduation from motivational data and reading achievement scores. Additionally, expectancy v alue theory will be reviewed and previous studies concerning motivational factors, reading achievement, and educational attainment will be discussed. Problem of Achievement Gap The term achievement gap, often cited in both professional and nonprofessi onal literature, essentially attempts to describe differences in achievement between groups of people, most notably among students of varying gender, ethnicities, and socioeconomic status (SES). Previous studies have shown that boys are falling behind the ir female counterparts in many subject areas, most significantly in reading and writing (NCES, 2007; Planty et al., 2008). The gender gap appears to be greater for students from low socioeconomic backgrounds and in particular for Latino and African Americ an males (Planty et al., 2008). The gender and cultural achievement gap, as well as the gap between lower and higher socioeconomic status groups, has been documented for many decades in the United States (Childrens Defense Fund [CDF], 2005; Freeman, 2004; NCES, 2007; Planty et al., 2008). Additionally, gender, cultural, and SES differences in achievement have been shown to exist internationally (Berliner, 2006; Francis & Skelton, 2005; OECD, 2003; West & Pennell, 2003). Clearly, schools
32 are not meeting the needs of millions of male students each year, and more often these boys are students of color and come from lower socioeconomic backgrounds (Bemak & Chung, 2005; Cholewa & West Olatunji, 2008; Education Trust, 2005a, 2005b; KewalRamani et al., 2007; Pl anty et al., 2008). First, data about the general achievement gap between males and females will be presented, then literature specific to the achievement gap between Latino students and White students will be discussed, followed by a presentation about i nformation about the impact of poverty on academic success. Gender Achievement G ap For several decades, through the National Center for Education Statistics, the U.S. Department of Education has tracked educational and demographic data on a wide range of academic constructs for elementary, middle, and high school age groups. Scores from the primary cognitive assessment used by the NCES, the National Assessment of Educational Progress survey (NAEP), have consistently shown the differences in achievement b etween males and females, most evident in the subjects of reading and writing (Planty et al., 2008). For example, for each grade level surveyed, a significantly higher percentage of females than males performed at or above the Basic, Proficient, and Advanced levels in reading, while more males than females annually performed below the Basic achievement level (Planty et al., 2008). Furthermore, in 4th, 8th and 10th grades, female students demonstrated significantly higher average writing scale scores than did their male peers for 2007 (Planty et al., 2008). As a whole, school aged males are clearly not performing as well as their female classmates in reading and writing.
33 Gender differences in educational attainment also exist. For example, from 2007 to 20 17, NCES data projects that men are expected to represent only 43% of total undergraduate enrollment, indicating that women are going on to pursue undergraduate degrees at a significantly higher rate than males (Adebayo, 2008; Peter, Horn, & Carroll, 2005; Planty et al., 2008). This gender gap in educational attainment is evidenced across all ethnic groups, with females earning close to two thirds of the postsecondary degrees awarded to African American, Latino, and Native American students (Blackhurst & A uger, 2008). While educational achievement and attainment levels demonstrate students abilities to perform well on standardized tests and make passing grades, they do not measure student effort or attitudes toward academics. One NCES national longitudi nal study revealed that male students repeatedly reported schoolwork to be less meaningful on average than did the females (Freeman, 2004). When asked how interesting students found most courses, males answered quite or very interesting less often than did the females. Similarly, female students reported feeling that their school learning would be quite or very important consistently more than did the males (Freeman, 2004). Female twelfth graders also reported that they often or always tried to do their best work on average more often than their male counterparts. When asked how often they fooled around in class, males responded often or always more frequently than did their female classmates (Freeman, 2004; NCES 2002). Furthermore, literature h as often focused on the prevalence of and negative effects of boys attitudes that success in school is not macho or cool (Francis, 2000, Francis & Skelton, 2005; Sax, 2007), and that studying is not as socially acceptable for them (Van Houtte, 2004). In
34 particular, as Latino boys progress through school, they tend to view lower achieving male peers more positively than higher achieving males, indicating a possible social acceptability of lower achievement for these culturally diverse boys (Taylor & Gr aham, 2007). Males are not only significantly behind in test scores and graduation rates. As a group, they also report fewer achievement related behaviors and less classroom compliance than their female counterparts, which can often serve as indicators of motivation for school success. Boys more often come to school unprepared, as measured by bringing materials and homework to class (Jianzhong, 2006; NCES, 2007). Additionally, males on average fail to complete or hand in assignments more often than fem ales, although there has been a slight improvement for the males in the frequency of turning in work over time (NCES, 2002). Girls are overrepresented in the advanced level courses (especially English classes), are more often the valedictorians, and have fewer discipline referrals than do their male counterparts (Clark et al, 2006; NCES, 2005). In fact, boys (and more often culturally diverse boys) are referred to the office and receive disciplinary consequences over four times as often as girls (Skiba, M ichael, Nardo, & Peterson, 2002), and are more likely to be viewed as problem students (Lopez, 2003). Boys also make up more than twothirds of the special education referrals in the United States (McIntyre & Tong, 1998; Skarbrevik, 2002; Smith, 2007; W ehmeyer & Schwartz, 2001). Males are more likely to be diagnosed with Attention Deficit Hyperactivity Disorder than females, which may result from gender differences in the presentation of behaviors associated with this diagnosis (Sciutto, Nolfi, & Bluhm, 2004). Males are almost twice as likely as females to be retained in a
35 grade and are much more likely to drop out of school completely (Lopez, 2003; NCES, 2006). Cultural A chievement G ap Data from the U. S. Department of Education and professional scho ol counseling and education literature indicates a continuing achievement gap between White students and students of color in most subject areas (Bemak & Chung, 2005; Education Trust, 2005a, 2005b; Howard, 2003; KewalRamani et al., 2007; NCES, 2007; Planty et al., 2008). In comparison with White students, Hispanic students more often performed below the Basic level in across grade levels (NCES, 2007). Additionally, the percentage of culturally diverse students scoring at or above Basic is significantly below that of their White peers across grade levels (Howard, 2003; NCES, 2007). These gaps in achievement are especially pronounced in the subject of reading ( see table 2 1, NCES, 2007). On the 2007 national reading achievement tests, Hispanic students averaged 25 points lower than White students (Planty et al., 2008). These cultural gaps in reading achievement scores have held relatively constant since 1992 (Llagas & Snyder, 2003; Planty et al., 2008). Similar cultural gaps have existed since the early 1990s in mathematics as well, with Hispanic students averaging scores far below White students at the 4th and 8th grade levels on the NAEP (Planty et al., 2008). White students also outperformed their Hispanic counterparts in writing achievement scores acr oss all grade levels (Planty et al., 2008). The previously discussed gender achievement gaps are especially pronounced for culturally diverse students (Chatterji, 2006; Lopez, 2003; Meece et al., 2006). Gender differences in motivation may also exist wit hin ethnic groups, with research indicating that Latina and African American high school girls report being more
36 motivated by homework, understanding class material, and gaining knowledge to help them in the future as compared to their respective male peer s (Gordon Rouse & Austin, 2002). Table 2 1. Percentage distribution of students across reading achievement levels, by race/ethnicity and grade: 2007 Grade and level Total 1 White Black Hispanic Asian/ Pacific Islander American Indian/ Alaska Native 4th grade Below basic 33 22 54 50 23 51 At or above basic 67 78 46 50 77 49 At or above proficient 33 43 14 17 46 18 At advanced 8 11 2 3 15 4 8th grade Below basic 26 16 45 42 20 44 At or above basic 74 84 55 58 80 56 At or above proficient 31 40 13 15 41 18 At advanced 3 4 # 1 5 2 # Rounds to zero. 1 Total includes races/ethnicities not separately shown. NOTE: Detail may not sum to totals because of rounding. The reading data include students for whom accommodations were permitted. For a discussion of the reading achievement level definitions, see http://nces.ed.gov/nationsreportcard/reading/achieve.asp Race categories exclude persons of Hispanic ethnicity. SOURCE: U.S. Department of Education, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2007 Reading Assessments and NAEP Data Explorer, retrieved January 25, 2008, from nces.ed.gov/nationsreportcard/nde/ Dropout and retention r ates As previously discussed, achievement gaps between populations are often demonstrated through outcome measures other than test score data. For example, the cultural achievement gap can also be seen in retention and dropout rates. In 2006, Hispanic students dropped out at an alarming rate of 22% in comparison to Black students (11%) and White students, who dropped out at a rate of 6% (Planty et al., 2008). While some educators c ite students immigrant status as an explanation for this significant difference in dropout rates, even when one looks only at Americanborn
37 children, Hispanic students have a higher dropout rate than any other ethnic group (Llagas & Snyder, 2003). In ad dition to differences in achievement levels, culturally diverse students are often overrepresented in areas of special education (MacMillan, Gresham, Lopez, & Bocian, 1996). Research suggests differences in gender and cultural expressions and classroom ex pectations lead to misunderstandings and inappropriate referrals for special education services (MacMillan et al., 1996; McIntyre & Tong, 1998; Neal, McCray, Webb Johnson, Bridgest, 2003). Furthermore, Latinos are greatly underrepresented in services for gifted students (Howard, 2003; Smith, 2007). Latino students are also less likely than White students (and equally as likely as other multicultural students) to enroll in and complete advanced level courses in high school (Llagas & Snyder, 2003). For example, an analysis of transcript data from the ELS:2002 study revealed that only 16% of Hispanic high school graduates in 2004 earned credit for an advanced placement or International Baccalaureate class, compared to a third of White students and a quarter of Black students (Planty, Bozick, & Ingels, 2006). Additionally, Latino students are more likely than White students to be expelled or suspended from school at some point during their 7th to 12th grade school years (Llagas & Snyder, 2003). Postsecondary enrollment and a ttainment Beyond high school, a cultural achievement gap persists in post secondary educational attainment (Saenz & Ponjuan, 2009). Significant cultural differences in the rates of immediate college enrollment (enrolling in college immediately after completing high school) are evident throughout recent decades: White students enrolled at 69%, African American students enrolled at 55%, and Hispanic students enrolled at a rate of
38 58% in 2006, and these differences have fluctuated little over time (Planty et al., 2008). Additionally, gender differences in the educational attainment across ethnic groups remain significant, with a greater percentage of females of each ethnicity attaining a bachelors degree or higher in 2007 (Blackhurst & Auger 2008; Planty et al., 2008). For example, 32% of White males between the ages of 25 and 29 earned at least a bachelors degree, while 19% of African American males and only 9% of Latino males earned this degree (Planty et al., 2008). In contrast, 40% of White females, 20% of African American females, and 15% of Latinas earned at least bachelors degree (Planty et al., 2008). According to Goldenberg, Gallimore, Reese, & Garnier (2001): A variety of factors have been nominated and investigated as contributing to the disproportionate underachievement of many ethnic and racial minority groups in our schools, e.g., poverty; cultural and linguistic discontinuities between home and school; the hidden curriculum of the classroom that privileges EuroAmerican, mi ddle class experiences; discrimination; and low aspirations or expectations rooted in inequalities and discrimination. (p. 575) Clearly, a cultural gap persists in educational attainment as well as academic achievement, with Latino students attaining high school and postsecondary degrees much less frequently than their nonHispanic peers. Socioeconomic Achievement G ap In addition to gender and cultural achievement gaps, differences in achievement between low and high SES students are numerous and substantial, and are well documented in the literature (Amatea & West Olatunji, 2007b; Berliner, 2006; Blair, Blair, & Madamba, 1999; Chatterji, 2006; Duncan & Brooks Gunn, 2001; Flowers & Flowers, 2008; Jones & Myhill, 2004; Pagani, Boulerice, Vitaro, & Tremblay, 1999; Smith, 2007; Varlas, 2005). In addition, achievement gaps across socioeconomic status levels exist for all ethnic groups, including Latino students (Berliner, 2006; Ma,
39 2000). Flowers and Flowers (2008) found that family income had a positive effect on the reading achievement of culturally diverse high school students. Family poverty also decreases students abilities to compete academically with their peers, predicts academic failure, and predicts students being in grade levels below their age group (Pagani et al., 1999). Literature has also demonstrated a frequent interconnection between SES and ethnicity: while the majority of children in poverty in the United States are White, roughly 60% of Latino and African American students come from famili es situated near or below the poverty line (Berliner, 2006; Planty et al., 2008; Smith, 2007). In 2006, approximately 26% of Hispanic students were living below the poverty line, in comparison to only 10% of White students (Planty et al., 2008). Signific antly larger percentages of culturally diverse students than White students attend high poverty schools, which are defined as schools where over 75% of students qualify for free or reduced lunch (Planty et al., 2008). In particular, roughly one third of H ispanic students attend a high poverty school, and almost 60% of Hispanic children attend schools where over 50% of the students qualify for free or reduced lunch programs (Llagas & Snyder, 2003). Additionally, professional literature points to the connection between poverty and gender: for reading in particular, the achievement gap between boys and girls is even wider for economically disadvantaged students than for middle class students (Chatterji, 2006; Francis & Skelton, 2005). Boys of lower socioeco nomic status perform less well than lower income girls on achievement tests and exams, nationally and internationally (Francis & Skelton, 2005; Planty et al., 2008). Additionally, researchers have found that
40 low income males may be perceived less favorabl y by their teachers when compared to low income females or higher income males (Auwarter & Aruguete, 2008). Many students from low SES families start school disadvantaged due to a lack of prenatal care, health insurance, immunizations, and good nutrition ( Berliner, 2006; CDF, 2005; Chatterji, 2006; Gutman, Sameroff, & Cole, 2003). Woolley and his colleagues (2008) also found that poor physical conditions associated with lower SES contributed to lower achievement in reading and mathematics, and that the poor conditions had more of an effect on achievement as students progressed through school. In addition, students from low income families may be at risk of lower achievement because of limited exposure to environmental stimulation at home such as books, cul tural events, and scientific or other costly enrichment activities (Chatterji, 2006; Pagani et al., 1999). For example, gaps in reading achievement have been shown to increase over the summer months as wealthier students have greater access to educational materials and experiences between school years when compared to students in poverty (Berliner, 2006; Cooper, Nye, Charlton, Lindsay, & Greathouse, 1996). High poverty areas of the country routinely lack high quality teachers and counselors, and have inappropriate funding for educational institutions and materials such as computers, library books, and supplemental enrichment programs (Amatea & West Olatunji, 2007b; CDF, 2005; Smith, 2007). In fact, a report by the Education Trust (2005) found an average funding gap between low and highpoverty schools and schools with higher proportions of culturally diverse students of over $1300 per student. Additionally, teachers in highpoverty schools turn over at greater rates than those in low poverty schools (Pl anty et al., 2008), leading to less continuity and stability in these
41 higher poverty schools. Furthermore, more of the nations students in poverty are located in urban settings, which presents an additional set of stressors that can affect achievement (B erliner, 2006; Lopez, 2003). Thus, even if students themselves are not in poverty, the quality of their education may be negatively impacted if they live in a high poverty area or attend a highpoverty school (Berliner, 2006; The Education Trust, 2005b). Teachers expectations also vary based on students socioeconomic status. Teachers are also more likely to rate low income males as less academically capable, and to have lower educational expectations for these young men when compared to both females and higher income males (Auwarter & Aruguete, 2008; Diamond, Randolph, & Spillane, 2004). Furthermore, when teachers rate low SES boys less favorably than highSES boys but rate low SES girls more favorably than highSES girls, this indicates an interactio n effect of gender and SES on teacher perceptions of ability and academic expectations (Auwarter & Aruguete, 2008). In other words, the low income boys might be particularly impacted by negative teacher expectations and evaluations (Auwarter & Aruguete, 2 008). Finally, teachers may believe that student SES determines academic success, and may feel ineffective or lack the motivation to intervene because SES is out of their control, which may further perpetuate achievement issues with low SES students (Auwarter & Aruguete, 2008). Beyond the school grounds, students raised in low income neighborhoods may lack successful academic role models in the community to serve as visible, motivating examples of achievement (Berliner, 2006). In fact, research suggests that the socioeconomic environment plays a stronger role than genetics in the expression of
42 intelligence for individuals living in poverty than for wealthier individuals, resulting in a misleading perception that all students from low SES backgrounds are unintelligent or unmotivated (Berliner, 2006; Duncan & Brooks Gunn, 2001; Turkheimer, Haley, Waldron, DOnofrio, & Gottesman, 2003). Clearly, school counselors need to be aware of how the environment of poverty profoundly impacts students cognitive devel opment and academic performance, and perhaps also the expression of motivation for these students. Like the gender and cultural achievement gaps, the difference in achievement between higher and lower SES students is demonstrated in educational attainment data. ELS:2002 data reveals that 90% of the 10th grade dropouts for the 20022003 school year were from the bottom 3 quartiles of socioeconomic status (HampdenThompson et al., 2009). Since the 1970s, the difference between low and high income student s in immediate college (2 or 4 year institutions) enrollment has generally decreased, although a significant gap still remains (Planty et al., 2008). Specifically, in 2006, 51% of low income students (lowest 20% of family incomes), 61% of middle income st udents (middle 60% of family incomes), and 81% of high income students (upper 20% of family incomes) enrolled in college immediately after completing high school (Planty et al., 2008). Among a range of factors, student socioeconomic status represents the strongest predictor of academic performance for both culturally diverse groups and White students (Berliner, 2006; Blair et al., 1999). As a result, the influence of poverty on childrens education represents the single biggest obstacle to achievement for students (Berliner, 2006; Jones & Myhill, 2004; Lee, 2005; Smith, 2007; Varlas, 2005).
43 Therefore, in their leadership positions school counselors represent key players in raising awareness and in improving the educational experiences of low income stude nts, and need to be aware of the significant impact of poverty on student achievement, motivation, and graduation (Amatea & West Olatunji, 2007b; Lee, 2005; Stone & Dahir, 2006). Masculinity and Intersectionality Connells (1987; 1995) concept of subtypes of masculinities informs the interaction of gender, culture, and environment with achievement in school. Connell (1995) proposes that ones gender identity (masculine or feminine) is developed over time through a n ongoing process of interaction with ones family and peers, and is set in the contexts of society, culture, school, and history. Given this sociocultural view of the development of masculinity, Connell (1995) describes hegemonic masculinity as the prev ailing or currently accepted expression of manhood in a society. In the United States, hegemonic masculinity is seen as White, middleto upper class, straight men (Connell, 1995; Hurtado & Sinha, 2008). Automatically based on this image, culturally diver se men, as well as men from low SES backgrounds and homosexual men, are devalued and seen as less masculine (Connell, 1987; 1995). These groups are marginalized because of their contrast to, or variance from, the White hegemonic masculinity (Connell, 19 95). Because Latino high school males also experience these pressures and stigmas, it is important to recognize the range of expressions of masculinity (Blair et al., 1999; Hurtado & Sinha, 2008). Latino students also point to the presence of a tradition ally Hispanic gender role pressure, machismo that encourages young Latino men to be strong and tough, and that could impact the stance these young men take toward their education (Saenz & Ponjuan, 2009; The College Board, 2010).
44 There exists a need for i ncreased understanding about diverse masculinities and for an exploration of the ways in which Latino boys from various socioeconomic groups experience achievement. When discussing the experiences of any population, and particularly of disenfranchised po pulations within a society, one must consider the combined influences of multiple identities, or the intersection of identities. For example, identities can be cultural (Latino, AfricanAmerican, Asian), socioeconomic (low SES, high SES), or gendered (male, female, gay, lesbian), among others. Shields (2008) emphasizes the need for research focusing on the intersection of identities as our society becomes more diverse. Moreover, the various identities to which a student belongs may interact to create unique influences on achievement. Blair et al. (1999) suggest that ethnicity can be envisioned as a quality whose effects within the educational area are difficult, if not impossible, to separate from the class based traits with which it is interwoven (p. 553). Lopez (2003) discusses the cumulative impact of stigmatizing racial and gender processes on Latino students at both the micro (student teacher interactions) and macro (system wide problems at low income or immigrant schools). For example, at the m icro level culturally diverse males in many urban schools are singled out for security checks and stigmatized as being troublesome or problematic, and are often disciplined more often than their female or White counterparts (Lopez, 2003, Skiba et al., 20 02). Additionally, at the macro level, the overwhelming lack of adequate resources and supplies combined with a less than challenging curriculum can lead to multicultural and low SES students to feel that their work and educational efforts are devalued, w hich can greatly impact achievement motivation (Lopez, 2003). While Blair
45 et al. discuss the intersection of ethnicity and socioeconomic status, Entwisle and her colleagues found that the intersectionality of being male, culturally diverse, and low SES pr edicts reading achievement (Entwisle, Alexander, & Olson, 2007; Hurtado & Sinha, 2008). Clearly, current research on academic achievement and motivation needs to consider the complex contributions of gender, ethnicity, and socioeconomic status to understand Latino males pathways to success. Previous A pproa ches to Addressing Achievement G aps Given the overwhelming statistics concerning male students (and Latinos in particular) and students in poverty who are underachieving and overrepresented in special education classes, research demonstrates that current educational practices may limit the success of millions of students (Cholewa & West Olatunji, 2008; Haycock, 2006). After No Child Left Behind (2001), states and school systems have been held accountable for the existing achievement gaps, but data shows that the efforts to close these gaps have had mixed results (Berliner, 2006; The Education Trust, 2005a). For example, some improvements have been demonstrated at the elementary school levels but not at the secondary levels, and in some states the achievement gap in reading has narrowed, but only because the achievement of White students has declined (The Education Trust, 2005a). Current efforts are not working, and this may be because programs have bee n implemented without input from teachers, counselors, and parents. Since many reform efforts are from the top down and are developed without significant school and community input, they often fail (Ginsburg, Shapiro, & Brown, 2004, p. xix). Perhaps another reason for the limited successes of current educational practices is because targeted populations have been historically viewed as monolithic groups, instead of recognizing the variance within groups and the significance of the intersection
46 of ident ities (Berliner, 2006). School counselors and other educators have identified the continued need for conversation about the complex impact of race, gender, and socioeconomic status on academic achievement in school (Dimmitt, 2003, p. 346). Therefore, w hen studying differences in achievement, there is a need to disaggregate student data into as many combinations of subsets to understand the dynamic relationships that exist within and between groups (Carpenter, Ramirez, & Severn, 2006, p. 123). There i s a need to more clearly understand the various patterns of achievement motivation of Latino male students from a range of socioeconomic backgrounds. Expectancy Value Theoretical Framework Eccles and her colleagues (1983) developed and tested a social cognitive expectancy value model of achievement related choices, expanding upon the expectancy value model proposed by Atkinson (1964). In the original Expectancy Value Theory, Atkinson proposed that the strength of the achievement motive (or tendency to ac hieve) is derived from the sum of a persons tendency to approach success and that persons tendency to avoid failure (Atkinson, 1957; Spence & Helmreich, 1983). Furthermore, the strength of these two opposing tendencies is determined by the motive to approach success or failure, the probability (or expectancy) of an achievement oriented behavior to result in success or failure, and the incentive value (task value) of success or failure (Spence & Helmreich, 1983). In subsequent years, other researchers have proposed additions and have elaborated on Atkinsons Expectancy Value Theory, choosing to focus more on cognitive and attribution processes that operate in real life situations instead of a laboratory (Spence & Helmreich, 1983).
47 The Eccles et al. (1983) Expectancy Value model represents a key example of this expansion of Atkinsons theory, in that it elaborates on several components of the original theory and incorporates a wider range of sociocultural and psychological components (Eccles & Wigfield, 2000, 2002). The general premise of this expanded theory is that individuals motivation for a task is influenced by expectancy related and taskvalue beliefs, and that these beliefs directly impact individuals performance, persistence, and task choices (Eccles et al., 1983; Wigfield & Eccles, 2000, 2002). Eccles et al. also include a socialization component that recognizes the role that the school (teachers, school counselors, and administrators), parents, and culture play in shaping student achievement beliefs and identity development (Meece, Glienke, & Burg, 2006), echoing the emphasis on context from ecosystemic perspectives (Bronfenbrenner, 1986). Below, the expectancy and task value components of this theory are explored in depth. Expectancies In this model, expectancies are defined as personal beliefs about how well one will do on a given task in the near or distant future (Eccles et al., 1983; Wigfield & Eccles, 2000, 2002). Furthermore, Eccles et al. propose that expectancies are influenced by self concept of ability, perception of task difficulty, perceptions of others expectations, causal attributions, and locus of control (Eccles et al., 1983; Wigfield & Eccles, 2000, 2002). Self concept of ability is defined as the assessment of ones own c ompetency to perform specific tasks or to carry out roleappropriate behaviors (Eccles et al., 1983, p. 82). Self concept of ability is also referred to as competency beliefs in existing literature (Meece et al., 2006). Longitudinal research has shown t hat students perceptions of
48 academic self competence decline not only after elementary school, but also through middle school and high school (Jacobs et al., 2004). Gender differences in competency beliefs have been shown to exist as early as elementary school, and many fluctuate through middle school years and persist into high school years as well (Meece et al., 2006). For example, girls begin school with roughly similar (if not slightly more positive) perceptions of their language arts abilities as compared to boys, but this gender gap increases as students move through middle school into high school, with girls having significantly higher perceptions of language arts abilities in middle school and high school (Marsh & Yeung, 1998; Meece et al., 2006). Often, these perceptions of competence persist even when students perform equally well in these subject areas (Meece et al., 2006). Evidence also suggests that boys perceptions of their competence decline more sharply than girls throughout schooling, in both math and language arts (Jacobs et al., 2004). Interestingly, Howard (2003) found that culturally diverse high school students connect perceptions of their academic identities and competence to their gender and ethnicity. Counseling literature has drawn attention to the cultural mismatch, or discontinuity, that students of color may experience in schools when their culture is not reflected in the classroom experiences and expectations (Cholewa & West Olatunji, 2008). Consequently, in efforts to ac hieve academic success, Latino students may feel that their school and cultural identities are at odds, which may present additional challenges to their motivation for success (Howard, 2003; Lopez, 2003). Finally, students self concepts of ability, espec ially when supported by parents and teachers,
49 are strongly linked to academic achievement and motivation (Cokley, 2003; Wigfield & Eccles, 2000, 2002; Gordon, 1995). Eccles et al. (1983) also suggest that gender differences in perceptions of task difficu lty may exist, and may impact academic achievement behaviors by influencing self concept of ability. While they admit that the effect of perception of task difficulty on achievement related behaviors is not clear, Eccles and her colleagues hypothesize that task difficulty may influence ones self concept of ability and may mediate achievement expectancies (Eccles et al., 1983; Wigfield & Eccles, 2000, 2002). While the importance of parental or teacher expectations and attitudes for student achievement has been researched and reported, Eccles et al. posit that the effect of these attitudes is mediated by the students perceptions of parent and teacher expectations (1983). It has also been well documented that teachers perceptions of students vary based o n student characteristics such as gender, ethnicity, and social class (Auwarter & Aruguete, 2008; Diamond, Randolph, & Spillane, 2004), and that these perceptions have an impact on student effort and achievement (Howard, 2003). Teachers more often rate boys of all ages as less competent in reading and writing than girls (Tournaki, 2003). Furthermore, Palardy (1998) reported that teachers beliefs about boys and girls literacy abilities impacted boys achievement on the Stanford Achievement Test at young ages: when teachers held negative expectations for boys, they performed significantly worse than their female peers. Additionally, teachers more often report lower academic expectations and abilities for low income students and for culturally diverse students than for students from wealthier families or from other ethnic backgrounds (Diamond et al., 2004). The teachers with lower expectations for low -
50 income and students of color also reported a decreased sense of responsibility for the students learning processes, which often perpetuates academic disadvantages of these populations (Diamond et al., 2004). Over time, these teacher perceptions are communicated to students through repeated verbal and nonverbal interactions, students begin to see themselves as their teachers see them, and their performance may then begin to match these perceptions (Howard, 2003; Palardy, 1998). Similar to teacher expectations, parent expectations vary depending on student gender and can influence students beliefs about their academic abilities (Meece et al., 2006). Parents play a key role in the development of student values and interests not only by supporting or discouraging certain activities or pursuits for their children, but also by serving as role models and exempl ifying the characteristics and skills that they think are important (Meece et al., 2006). High parental expectations for student educational attainment positively influenced the reading achievement (Flowers & Flowers, 2008), grade point averages (Somers, Owens & Piliawsky, 2008), and the educational aspirations (Flowers, Milner, & Moore, 2003; Howard, 2003) of multicultural high school students. Furthermore, Goldenberg et al. (2001) discredit the possible perception that parents of Latino students have low academic aspirations for their students, and instead found generally high levels of aspirations as expressed by the Latino parents in their longitudinal study. Additionally, the longer that parents of Latino students have been in the United States, the more likely they are to expect that their children go to college (Goldenberg et al., 2001). Research also shows that students of color who have expectations to complete high school also tend to have better grades,
51 suggesting that a future orientation can motivate students to achieve (Somers, Owens, & Piliawsky, 2008). Eccles et al. (1983) also suggest that whether a student attributes success or failure to stable or unstable factors influences that students expectancies for success. Weiners attribution theory (1972) provides the foundation for this aspect of expectancy value theory. Certain types of attributions have been shown to predict differences in motivation and achievement outcomes (Perry, Stupnisky, Daniels, & Haynes, 2008). Banks and Woolfso n (2008) found that students who see themselves as low achievers report less perceived control over poor school performance than students who saw themselves as higher achievers, indicating the connection between perceptions of competence and causal attribu tions. Finally, a students locus of control represents the extent to which a student perceives having control over their life or in a given situation (Eccles et al., 1983). Somers et al. (2008) found that belief in personal control was related to posit ive educational outcomes for culturally diverse high school students, and that this personal control may partially explain how educational intentions correlate with academic success. Additionally, Flowers, Milner, & Moore (2003) reported that high school students of color with higher levels of locus of control (beliefs in personal control) had higher academic aspirations as well. A greater sense of control was also found to differentiate more resilient and higher achieving culturally diverse students from lower achieving peers (Gordon, 1995). Finally, Sciarra & Whitson (2007) noted that locus of control and self efficacy are important variables to consider and include when researching the educational attainment of Latino youth.
52 Task values Task value is the importance a person places on succeeding or failing at a certain task (Eccles et al., 1983, Wigfield & Eccles, 2000, 2002). The task value is determined by the attainment value of the task, the intrinsic or interest value of the task, and the utility value of that task for achieving a future goal (Eccles et al., 1983; Wigfield & Eccles, 2000, 2002; Meece et al., 2006). Task values are influenced by cultural contexts and societal experiences, expectations, and influences (Taylor & Graham, 2007). Furthermore, researchers have found that task values (and gender differences in these values) are domain specific, or unique to a subject or task, and not universal (Jacobs et al., 2004). As with expectancy beliefs, gender differences in task value have been shown at various age groups, with girls valuing language arts more than boys across all school years (Meece et al., 2006). Interestingly, this difference in valuing of language arts persists even though value perceptions declined overall for both boys and girls over the course of schooling (Jacobs et al., 2004; Meece et al., 2006). Because measures of attitudes of task value are subject to social desirability bias, other authors have attempted to understand task value through students selec tion of admired peers (Graham, Taylor, & Hudley, 1998; Taylor & Graham, 2007). This line of research has demonstrated that 7th grade Latina and African American girls and White boys will rate higher achieving peers as more favorable and admirable, whereas Latino and African American boys will rate lower achieving peers as more respected, perhaps indicating the value that these students place on academic achievement (Graham, Taylor, & Hudley, 1998; Taylor & Graham, 2007). It is important to note that in early elementary grades, Latino, African American, and White boys and girls all rated higher achieving
53 peers as more admirable, which may suggest a differential socialization process that occurs as ethnic minority boys enter adolescence (Taylor & Graham, 200 7). Attainment value, described by Eccles and her colleagues (1983) as the importance a person places on doing well on a task, incorporates ones perception of the tasks ability to confirm salient and valued characteristics of the self (e.g., masculini ty, femininity, competence), to provide a challenge, and to offer a forum for fulfilling achievement, power, and social needs (p. 89). If a student sees himself as a skilled writer, he may be more inclined to register for a higher level English course that could offer a challenge and a chance to prove his competence in this subject area. Intrinsic value is described simply as a persons interest and enjoyment in participating in an activity or task (Eccles et al., 1983; Wigfield & Eccles, 2000, 2002). Flowers (2003) found that culturally diverse high school students who read outside of school for enjoyment performed better on standardized reading achievement tests than their peers who did not read for pleasure. Another study demonstrated that teachers hold stereotypes that boys as a group are disinterested in writing and are not skilled readers (Jones & Myhill, 2004). Researchers have also found that boys value the task of reading less than girls, and see reading as less pleasurable (Schwartz, 2002; Sm ith & Wilhelm, 2002). Finally, utility value is the usefulness of a given task for achieving some future goal (Eccles et al., 1983; Wigfield & Eccles, 2000, 2002). For example, a student may choose to take a certain class, not because they are interested in that class per se, but because taking that class allows them to pursue more interesting classes or activities in the future. Additionally, Eccles et al. (1983) suggest that the value of a task may be
54 mediated by sex roles, cultural experiences, perceptions of the cost of success (including effort, loss of valued alternatives, and psychological cost of failure), and affective experiences with similar tasks in the past. Gordon (1995) found that resilient students of color (those with better grades and higher academic motivation despite experiencing hardship) valued material gain more than their less resilient peers, and suggested that perhaps these students recognize the link between material gain and economic independence. In this chapter, literature r elevant to understanding the existing gender and socioeconomic achievement gaps for Latino students was reviewed, and expectancy value theory of motivation from Eccles and colleagues was presented as the guiding theoretical approach for this study. In the next chapter, the research design of this study as well as the national database to be used will be discussed. Finally, the proposed statistical analyses and procedures will be reviewed.
55 CHAPTER 3 METHODOLOGY The purpose of this study was to use Eccles et al. (1983) Expectancy Value Theory of motivation and tenth grade reading achievement scores to identify the factors that predict high school graduation for Latino males from various socioeconomic levels. The Education Longitudinal Study of 2002 (ELS:2002), a large, nationally representative, longitudinal database developed by the National Center for Education Statistics, was used to explore the relationship between aspects of motivation and reading achievemen t and graduation for Latino males. Specifically, this study addressed the following research questions: how do motivational factors and tenth grade reading achievement scores predict high school graduation for Hispanic males, and to what extent do these f actors differ among Hispanic males from varying levels of SES? In this chapter, the Educational Longitudinal Study of 2002 (the national database to be analyzed in this study), the study participants, the selected research variables, and data analysis procedures are discussed. Education Longitudinal Study of 2002 The Education Longitudinal Study of 2002 (ELS:2002) represents the most recent addition to a series of national longitudinal studies conducted by the U.S. Department of Educations National Cent er for Education Statistics (NCES) that began in 1972 with the National Longitudinal Study of the High School Class of 1972 (Ingels & Scott, 2004). This study, along with its predecessors, is designed to provide policymakers and researchers with longitudi nal and trend data about students experiences within the education system, their subsequent postsecondary education access and attainment, and their entry into the labor market (Ingels & Scott, 2004). The structure and content
56 of the ELS:2002 data is des igned to allow researchers to conduct studies at multiple levels, including cross sectional profiles, longitudinal analyses, intercohort comparisons with cohorts from previous NCES studies, and international comparisons (Ingels, Pratt, Rogers, Siegel, & St utts, 2004; Ingels & Scott, 2004). The ELS:2002 study baseyear elements included a survey of high school sophomores; cognitive tests in reading and mathematics; a survey for parents, English and mathematics teachers, and school administrators; a school f acilities checklist; and a library questionnaire (Ingels et al., 2004; Ingels & Scott, 2004). A two stage stratified probability sample design was used in the ELS:2002 study: the first stage of selection was schools, and the second stage of selection was students (Ingels et al., 2004; Ingels & Scott, 2004). The ELS:2002 baseyear study was conducted using a nationally representative sample of 752 public, Catholic, and private schools in the spring of the 2001 02 academic year (Ingels & Scott, 2004). Fro m these 752 schools, 743 principals and 718 librarians completed questionnaires (Ingels & Scott, 2004). Within the participating schools, 17,591 eligible 10th grade students were randomly selected to participate, and 15,362 actually completed a baseyear questionnaire (Ingels et al., 2004; Ingels & Scott, 2004). In addition to the student respondents, 13,488 parents and 7,135 teachers also completed questionnaires about themselves, the participating students, their interactions with the school, and their activities (Ingels & Scott, 2004). In the present study, the student is assumed to be the unit of analysis and data is taken from the student survey. Students answered questions about their background, school experiences and activities, future goals and
57 plans, employment, extracurricular activities, and attitudes toward education (Ingels & Scott, 2004). The ELS:2002, like other studies through NCES, incorporates a weighting scheme to account for the unequal probabilities of selection of schools and students in the baseyear sample, and to adjust for the unequal response rates of schools and students (Ingels & Scott, 2004). The student weight is a cross sectional weight that allows for researchers to generalize results to the population of questionnaireel igible 10th grade students in the 20012002 school year (Ingels & Scott, 2004). The purpose of creating these panel weights was to facilitate analyses designed to examine how the two student populations change over time, and can be used to conduct intracohort or cross cohort analyses (Ingels et al., 2007, p. 153). For example, in this study the student panel weight can be used to look at a cohort of sophomores in 2002 and determine the proportion of that cohort that graduated in 2004. The ELS:2002 als o uses design effects, which measure the impact of the departures of the ELS:2002 complex sample design from a simple random sample design on the precision of sample estimates (Ingels et al., 2007, p.143). For first follow up questionnaire data from all respondents, the average design effect for the ELS:2002 was 2.23, while the average design effect for the NELS:88 was 3.73, indicating that the ELS:2002 panel sampling was more efficient than the NELS:88 panel sampling for each respective sophomore class (Ingels et al., 2007). Additionally, when broken down into subpopulations (including dropouts), the ELS:2002 remained more efficient on average than the NELS:88 (Ingels et al., 2007).
58 Participants Of the participating students from the baseyear ELS:2002 cohort, half are male and half are female (Ingels & Scott, 2004). Additionally, 36% of the responding sophomores identify themselves as belonging to a culturally diverse group (Black/African American, Asian, American Indian, or Hispanic/Latino), 60% identified as White, and the remaining 4% identify themselves as having more than one cultural background (Ingels & Scott, 2004). Of the sophomores in the ELS:2002 base year study, Hispanic students comprise 16% of the respondents, representing the largest culturally diverse group among those surveyed (Ingels & Scott, 2004). During their sophomore year in 2002, most students (92%) reported that they attended public schools, 4% attended Catholic schools, and 3 % attended other private schools (Ingels & Scott, 2004). Additionally, of the baseyear respondents, 30% attended an urban school, 50% attended a suburban school, and 20% attended a rural school (Ingels & Scott, 2004). For this study, only Latino males were selected for analysis. In addition to demogra phic information about the participants, Ingels and Scott (2004) provide data on the educational expectations and the tested achievement levels of the 2002 sophomore class who participated in the ELS:2002 baseyear study. For example, 90% of students reported a definite expectation of their future educational attainment, and a large majority expressed a goal of attaining a college degree or higher (see Table 31). The ELS:2002 also provides data on students reading achievement levels as measured by their reading standardized score (T score) on the National Education Assessment Program, an examination that is administered in conjunction with the ELS:2002 survey (Ingels & Scott, 2004).
59 Table 3 1. Percentage of high school sophomores, by students highest level of education expected: 2002 Level of Education Percent Less than high school 0.9 High school completion or GED 7.3 Attend or complete 2 year community college or vocational school 6.4 Attend 4 year program, but not complete degree 3.9 Graduate from college 35.8 Obtain masters degree or equivalent 19.7 Obtain Ph.D., M.D., or other advanced degree 16.1 Dont know 9.8 NOTE: Details may not sum to totals because of rounding. SOURCE: U.S. Department of Education, National Center for Education Statistics, Education Longitudinal Study of 2002 (ELS:2002). Access Access to the ELS:2002 was obtained by following all NCES protocol relating to public use data access procedures as indicated on the study website ( http://nces.ed.gov/surveys/els2002/avail_data.asp). The public use data can be obtained in one of several ways, depending on the particular data needed. For public use data including only baseyear and first follow up surveys, interested parties can simply contact NCES through electronic mail for a CD copy of the public use data files and the Electronic Codebook (ECB), and this data will be mailed to them free of charge. For access to public use data that includes the second foll ow up information, which was needed for this study, data can be downloaded to a personal computer using a new Education Data Analysis Tool (EDAT) computer application by accessing the EDAT website ( http://nces.ed.gov/edat/ ) and following the command prompts. Additional protocol exists to obtain access to ELS:2002 restricted use data files that contain personally identifiable information because of the sensitive and confidential nature of these data.
60 Research Variables Dependent Variable The dependent variable in this study was graduation from high school, and was a dichotomous variable based on whether or not a student graduated from high school on time. The original categories of responses for this variable, F2F1HSST, included fall 2003 summer 2004 graduate, prefall 2003 graduate, received certificate of attendance, received GED, no high school credential as of summer 2004, and graduated, unknown if by August 2004. For the purposes of this study, only those students who graduated ontime (fall 2003 summer 2004) or early (prefall 2003) were considered as graduating, so this variable was recoded into a new variable, Graduationyesno, with only two categories, non high school grad and high school grad. Independent Variables Graduation may be influenced by multiple factors relating to previous achievement and motivation. Independent variables for this study were selected based on Expectancy Value theory of Eccles and her colleagues (1983; Wigfield & Eccles, 2000). ELS:2002 s tudent questionnaire items were selected that were similar to the original Expectancy Value questionnaire (Eccles et al., 1983; Wigfield & Eccles, 2000) and that related conceptually to these theoretical constructs, an d these items were then examined using confirmatory factor analysis to determine the appropriateness of the factors. Additionally, several composite variables had already been created and included in the second follow up dataset, and these were included w hen appropriate. See Table 3 2 for a summary of the constructs and applicable questionnaire vari ables from the ELS:2002 study.
61 Table 32. Variable label, name, and description Variable Label Variable name Variable description Socioeconomic status BYSES2 Composite SES variable Reading achievement BYTXRSTD Standardized reading test score How far parent expects student to go in school BYP81 From the parent survey How far English teacher expects student to go in school BYTE20 From the teacher survey BY English self efficacy scale BYENGLSE Composite variable comprised of BYS89C, BYS89F, BYS89I, BYS89K, BYS89M BY control expectation scale BYCONEXP Composite variable comprised of BYS89E, BYS89N, BYS89Q, BYS89T Parents expect success in school BYS27I Ord inal variable Teachers expect success in school BYS27H Ordinal variable Student perception of school counselors desire for student after high school SCE Reverse coded from BYS66E by researcher How far in school student thinks he will get BYSTEXP Composite variable based on BYS56 BY Instrumental motivation/utility interest scale BYINSTMO Composite variable comprised of BYS89D, BYS89H, BYS89P Enjoys reading ENJOY Composite variable comprised of BYS87B, BYS87D, BYS87Ecreated by researcher Importance of good grades BYS37 Ordinal variable Importance of good education BYS54O Ordinal variable How much likes school BYS28 Ordinal variable Classes are interesting and challenging BYS27A Ordinal variable High school graduation status as of August 2004 GRADUATIONYESNO Dichotomous variable created from F2F1HSST; dependent variable Note: BY in variable label, name, and description categories is used to represent scores from the base year survey.
62 Expectancy variables Students expectancies for academic achievement are influenced by their self concept of ability, perception of task difficulty, perception of others expectations, causal attributions, and locus of control (Eccles et al., 1983). Students self concept of ability suggests their general sense of self worth and competence and were comprised of ELS:2002 questionnaire items such as How well do you do the following: understand spoken English, speak English, read English, write English, and I am confident I can do an excellent job on m y English assignments. Perceptions of task difficulty describe how hard a task will be to master, and were assessed by items such as Im certain I can understand the most difficult material presented in English texts. Perceptions of others expectatio ns are measured directly in the ELS:2002 baseyear student questionnaire, asking students about what key adult figures in their lives expect for them after high school. This construct included questions like How far in school do you think your mother and father want you to go? where students select from a range of postsecondary education options. Causal attributions are the extent to which students attribute success with stable or unstable factors, such as ability or luck. This construct was measured by questionnaire items such as If I want to learn something well, I can. Finally, locus of control represents the degree to which students feel they have control over a situation or a task. This construct was assessed by questionnaire items such as I a m certain I can master the skills in my English class. The ELS:2002 database included several composite variables created by NCES through principle factor analysis from multiple questionnaire items. For example, the BYENGLSE variable is a scale of stud ents base year English self efficacy, with higher scores indicating higher self efficacy. This composite variable was comprised of five
63 questionnaire items: Im certain I can understand the most difficult material presented in English texts, Im confi dent I can understand the most complex material presented by my English teacher, I am confident I can do an excellent job on my English assignments, Im confident I can do an excellent job on my English tests, and Im certain I can master the skills being taught in my English class. The BYENGLSE variable addresses the self concept of ability, the perception of task difficulty, and the locus of control constructs from Expectancy Value theory. Another composite variable created by NCES using variabl es from the ELS:2002 database was the BYCONEXP variable, which is a scale that measures respondents success expectations at the base year. Higher values represent higher expectations of success in academics. This variable was created using principle fac tor analysis and the base year student weight from four questionnaire items: When I sit myself down to learn something really hard, I can learn it, If I decide not to get any bad grades, I can really do it, If I decide not to get any problems wrong, I can really do it, and If I want to learn something well, I can. The BYCONEXP variable addresses the causal attributions construct within Expectancy Value theory. Task value variables Task value can be defined as the importance of success on a given activity. Task value is influenced by attainment value, intrinsic/interest value, and utility value, as well as cost of success (Eccles et al., 1983). Attainment value, the importance of succeeding on a task, was measured by questionnaire items such as How important are good grades to you? Intrinsic or interest value is the amount of enjoyment a person gains from participating in an activity, and was measured using questionnaire items such as how much students agree with the statement Because reading is fun, I
64 wouldnt want to give it up. Utility value, or the extent to which succeeding at a task will help a person attain some future goal, was measured using ELS:2002 questionnaire item such as students agreement with the statement I study to get a good job. Finally, cost of success is the amount of time or energy that is devoted to a task that takes focus away from other activities. For example, cost of success was measured by using items such as In your current English course, how much time do you spend on homework each week, both in and out of school? As with the NCES created composite variables that address student expectancies, the ELS:2002 also has a composite variable that addresses task values. The BYINSTMO variable measures base year respondents instrumental (extrinsic) motivation or utility value, with higher scores representing higher instrumental motivation. In Expectancy Value theory, this composite variable, in addition to other questionnaire items, addresses both the intrinsic/ interest value construct as well as the utility value component. The BYINSTMO variable was created using principle factor analysis and the base year student weight from three questionnaire items: I study to get a good job, I study to increase my job opportunities, and I study to ensure that my future will be financially secure. Reading achievement An additional independent variable in this study was reading achievement as measured by the reading standardized score (T score) on the National Education Assessment Program, administered as a part of the ELS:2002. This variable is coded as BYTXRSTD in the ELS:2002 database. This score represents a norm referenced of reading achievement. Reading scores are based on three levels of mastery:
65 Simple readi ng comprehension, including reproduction of detail and/or the authors main thought; (2) Ability to make relatively simple inferences beyond the authors main thought and/or understand and evaluate abstract concepts; and (3) Ability to make complex inferences or evaluative judgments that require piecing together multiple sources of information from the passage. (Ingels & Scott, 2004, p. 5) Data Analysis The data analysis for this research depended on determining the best operationalization of the construc ts used in the model. Before analyses could be conducted using the data from the ELS:2002, the raw data had to be conditioned and some variables needed to be recoded for ease of analysis. To condition the data, the computer program SPSS was used. Additi onally, confirmatory factor analysis was used to create composite variables from selected ELS:2002 questionnaire items, guided by constructs within expectancy value theory. Confirmatory factor analysis represents one statistical tool within the larger cat egory of structural equation modeling, or covariance structure analysis (Thompson, 2004). Factor analysis is used to determine which variables (or question items) clump together and correlate more strongly with each other than with other variables to create factors, or groups of closely related variables (Aron & Aron, 1999, Kline, 2005). Factors may also be defined as underlying latent variables, or variables that cannot be directly measured (Kline, 2005). The correlation of each individual variable (or question item) to a factor is its factor loading on that factor (Aron & Aron, 1999). The computer software program SPSS was used to conduct the factor analyses, and to generate a Cronbachs alpha level as a measure of internal consistency for these variables. Confirmatory factor analysis specifies how the measured variables reflect certain latent variables, and therefore represents a preliminary measurement step before conducting additional structural analyses
66 (Thompson, 2004, p. 110). In summary, conf irmatory factor analysis was used to determine the fit of items to the particular underlying factors. After using factor analysis to determine the latent factors comprised of related questionnaire items, a structural regression analysis was conducted first for the 10th grade Latino male population as a whole, and then for each SES quartile within this sample. To conduct these regression analyses, the MPLUS software program was used, which is designed to handle the complex sampling designs and weighting pro cedures like those in NCES databases (Kline, 2005). Structural regression, another tool within structural equation modeling, is ideal for testing models that have both a structural component (similar to path analysis) and a measurement component (similar to confirmatory factor analysis) in the design (Kline, 2005). Path analysis would not have been appropriate for this data because this method requires a single measure of each construct (Kline, 2005), but in the proposed model there were several indicator s for each latent construct from Expectancy Value theory. Instead, structural regression incorporates a measurement component that allows for the analysis of observed variables as indicators of latent variables, as in confirmatory factor analysis (Kline, 2005; Thompson, 2004). Furthermore, structural regression allows for the analysis of these latent variables (or factors) within the structural model, rather than being limited to single observed variables as required by path analysis (Kline, 2005). Howev er, like path analysis, structural regression also allows for the testing of hypotheses about direct and indirect effects (Kline, 2005). Finally, because graduation is a dichotomous variable, logistic regression was needed. For an analysis of each socioeconomic group, structural regression was used to estimate the model of indicators and factors that
67 significantly predict high school graduation and the models were examined and compared. This chapter presented information about the ELS:2002, its participants, and how researchers can access this database. Additionally, the selection of research variables based on expectancy value theory of motivation by Eccles and her colleagues was explored. Finally, the appropriate data analysis techniques and statisti cal procedures were discussed. The following chapter will present the results of the analyses proposed in this chapter, and Chapter 5 will present a discussion of these findings.
68 CHAPTER 4 RESULTS The purpose of this study was to use Eccles et al. (1983) Expectancy Value Theory of motivation and tenth grade reading achievement scores to identify the factors that predict high school graduation for Latino males from various socioeconomic levels. Spec ifically, this study addressed the following research questions: how do motivational factors and tenth grade reading achievement scores predict high school graduation for Hispanic males, and to what extent do these factors differ among Hispanic males from varying levels of SES? The previous chapter described the Education Longitudinal Study of 2002, the research variables that were selected, and the research methodology that was used in this study. In this chapter, the results of the previously discussed data analysis are presented. General descriptive statistics about the data will be provided, followed by a review of the results of the factor analysis and of the structural regression analysis. Descriptive Statistics To conduct the data analysis, the E LS:2002 dataset needed to be conditioned to prepare the data for further analyses. After selecting for only Hispanic males, the data was reduced to 1099 cases (or individuals). Descriptive statistics for all the variables used in this study were calculat ed using one of the panel weights for the data, BYSTUWT. The purpose of applying weights to student data is to account for the unequal probability of selection and sampling of certain populations. According to the ELS:2002 data documentation manual, this particular panel weight is applied to spring 2002 sophomores who completed data in the base year, and can be used for cross cohort comparisons, and allows results to generalize to all 10th grade students capable
69 of completing the questionnaire (Ingels et al., 2007). Applying the appropriate student weight to the narrowed population of interest (n=1099) allows projections to be made to all 10th grade high school Hispanic males in the United States capable of completing the questionnaire, resulting in a tot al number of weighted cases of 267,622 students. See table 41 for a summary of descriptive statistics for the variables in this study, and refer to table 42 for a summary of the correlations between the variables. Table 41. Descriptive Statistics for Variables in Model Variable Name Mean Standard Deviation Minimum Maximum BYS28 2.15 .613 1 3 BYENGLSE .141 .971 2.197 1.596 BYCONEXP .168 1.016 2.521 1.580 BYINSTMO .078 .970 1.994 1.579 GRAD .68 .467 0 1 BYS27A 2.33 .800 1 4 BYS27I 1.49 .706 1 4 BYS27H 2.27 .854 1 4 BYS37 3.26 .809 1 4 BYS54O 2.78 .457 1 3 BYP81 4.81 1.547 1 7 BYTE20 3.44 1.469 1 7 BYTXRSTD 44.326 9.699 22.69 71.98 ENJOY 0 1.00 1.446 .760 SCE 6.196 1.858 1 7 BYSTEXP 4.00 2.121 0 7 BYSES2QU 1.89 1.029 1 4 When compared to all 10th grade high school male and female students of all ethnic groups, 48.4% of the Latino males in this study were categorized as being within the lowest socioeconomic quartile, 24.2% were categorized in the second quartile, 17.0% in the third quartile, and 10.4% in the highest quartile. The mean quartile classification for the 10th grade Latino males was 1.89, with a standard deviation of 1.029.
70 Table 4.2. Correlation matrix for variables included in the model BYS28 BYENGLSE BY CONEXP BYINSTMO GRAD BYS27A BYS27I BYS27H BYS28 1 BYENGLSE 0.229 1 BYCONEXP 0.226 0.698 1 BYINSTMO 0.296 0.563 0.688 1 GRAD 0.183 0.233 0.269 0.214 1 BYS27A 0.505 0.146 0.116 0.232 0.069 1 BYS27I 0.22 0.235 0.252 0.219 0.273 0.175 1 BYS27H 0.433 0.097 0.168 0.193 0.151 0.508 0.514 1 BYS37 0.435 0.23 0.287 0.379 0.222 0.277 0.377 0.316 BYS54O 0.291 0.243 0.275 0.308 0.197 0.178 0.304 0.291 BYP81 0.207 0.173 0.215 0.203 0.218 0.12 0.158 0.076 BYTE20 0.136 0.265 0.253 0.222 0.506 0.138 0.129 0.089 BYTXRSTD 0.024 0.289 0.278 0.179 0.349 0.113 0.074 0.128 ENJOY 0.255 0.286 0.238 0.24 0.014 0.264 0.091 0.263 SCE 0.059 0.101 0.119 0.188 0.032 0.114 0.069 0.15 BYSTEXP 0.159 0.212 0.243 0.325 0.213 0.106 0.134 0.051 Table 4.2 (cont.). Correlation matrix for variables included in the model BYS37 BYS54O BYP81 BYTE20 BYTXRSTD ENJOY SCE BYSTEXP BYS37 1 BYS54O 0.477 1 BYP81 0.245 0.199 1 BYTE20 0.22 0.137 0.292 1 BYTXRSTD 0.07 0.078 0.22 0.467 1 ENJOY 0.18 0.12 0.136 0.042 0.073 1 SCE 0.13 0.054 0.152 0.125 0.026 0.009 1 BYSTEXP 0.259 0.208 0.256 0.281 0.261 0.119 0.204 1 Note: Correlations that are statistically significant at the p = .05 level are noted with an asterisk
71 For 10th grade Latino males in this study, 47.8% reported English as their native language, while 43.5% reported Spanish as their native language. As of the summer of 2004, which was two years after the base year questionnaire, 67.8% of the 10th grade Latino males in this study had graduated from high school and 32.2% of this sample had not received a high school diploma. When asked how much they like school, 60.3% of the 10th grade Latino males reported that they somewhat liked school, 12.5% reported that they liked school not at all, and 27.2% reported that they liked school a great deal. When asked to respond about the personal importance of good grades, 3.2% of the 10th grade Latino males selected that grades were not important to t hem, 13.7% reported that grades were somewhat important, 37.4% stated grades were important, and 45.8% reported that good grades were very important to them. Similarly, when asked about the importance of getting a good education, 1.9% of the 10th gr ade Hispanic male survey respondents reported a good education was not important, 17.8% reported it was somewhat important, and 80.3% reported a good education was very important to them. For the 10th grade Hispanic males in this study, the mean sc ore on BYENGLSE (baseyear English self efficacy composite scale score) was .141 with a standard deviation of .972, with a minimum of 2.197 and a maximum of 1.596. For the entire population of sophomores who completed the ELS:2002 base year survey, NCES standardized the mean for this variable at 0 with a standard deviation of 1. The mean score on BYCONEXP (baseyear control expectation composite scale score) for this sample was .168 with a standard deviation of 1.017, a minimum score of 2.521 and a
72 ma ximum score of 1.580. Again, NCES standardized the mean for all base year survey respondents at 0 with a standard deviation of 1. The mean score on BYINSTMO (baseyear instrumental motivation/utility interest composite scale score) for the 10th grade Lat ino males who responded to this survey was .078 and the standard deviation was .971, while the mean for all base year respondents was again standardized by NCES at 0 with a standard deviation of 1. The mean reading test standardized T score (BYTXRSTD) for the 10th grade Latino male respondent sample was 44.326, and the standard deviation was 9.703, with a minimum score of 22.69 and a maximum score of 71.98. For the entire base year respondent population, the mean reading test standardized T score was 50 w ith a standard deviation of 10, indicating that the mean reading score for 10th grade Latino males is less than that of the entire base year respondent population. Factor Analysis Results The researcher created one variable through factor analysis using the SPSS computer program. The variable, named ENJOY, combined three questionnaire items from the base year student survey. These three questions asked students to rate how much they agreed or disagreed with the statements Because reading is fun, I woul dnt want to give it up (BYS87B), I read in my spare time (BYS87D), and When I read, I sometimes get totally absorbed (BYS87E). The ENJOY variable was created using principal factor analysis weighted by the base year student weight (BYSTUWT), following the example of the ELS:2002created composite variables. Higher score on this variable indicate higher enjoyment of reading. The coefficient of reliability for this variable (Cronbachs alpha) was .973. The mean score on this variable for the 10th gr ade Latino males capable of completing this survey was standardized at 0 with a
73 standard deviation of 1, as this variable was created using just the sample described above. The minimum score on this variable for the 10th grade Latino males who completed th e survey was 1.446 and the maximum score was .760 for this same group. Structural Regression Analysis To conduct the data analysis for this study, the computer program MPLUS was used because of this programs ability to automatically account for the com plex analysis needed for the complex sampling design of the ELS:2002 database, and to allow the inclusion of all individuals (or cases) with any missing data (Kline, 2005). Additionally, the particular statistical method selected, structural regression, i s a type of structural equation modeling that accounts for complex sampling inherent in these types of national databases (Kline, 2005). To define the model for this study, the original model proposed in Chapter 3 was entered into MPLUS (see Figure 4 1), and the goodness of fit was assessed. The goodness of fit of any given model can be measured by several methods including the Chisquare test of model fit score, the Comparative Fit Index (CFI), and the Tucker Lewis Index (TLI). The Chi square tes t of model fit score is evaluated based on the associated p value to determine significance. The maximum CFI score is 1.000, and CFI scores of greater than .95, are generally considered to indicate good fit. Similarly, TLI scores greater than .95 also indicate good model fit. The root mean square error of approximation (RMSEA) compares the estimates for the variable means, variances, and covariances between variables under the assumption that the model is correct to those same estimates without this assumption. Lower RMSEA estimates indicate a better model fit. The goodness of fit tests for the proposed model based on Expectancy Value
74 Figure 41. Hypothesized model based on Expectancy Value Theory
75 theory resulted in a significant chi square s tatistic of 700.175 ( p = 0.000), indicating a significant lack of fit of the data to the model. The other goodness of fit indices also revealed a poor model fit (CFI = .774, TLI = .547, RMSEA = .099). Based on these results, an attempt was made by the researcher to modify the original model by adding parameters one at a time; however, this attempt resulted in models for which the estimat ion procedure would not converge. As a result, a full model specifying all potential predictive associations between variables and all correlational associations among residuals for pairs of dependent variables was then entered into MPLUS. This method o f model entry was used by the researcher to ensure that the MPLUS estimation procedure converged. The correlational associations in this full model allow for the residuals of the dependent variables to covary, and account for the possibility that the dependent variables share predictors that were not included in the model. On the other hand, predictive associations indicate a variables ability to predict a change in another variable. In other words, a statistically significant predictive association indicates that a change in one variable predicts a change in the subsequent variable. Entering this full model in the MPLUS program resulted in a statistically significant chisquare value of 0.000 ( p = 0 .000) indicating a nonsignificant lack of fit of th e data to the model and perfect goodness of fit indices (CFI = 1.00, TLI = 1.000, RMSEA = 0.000). Overall, the model exhibited excellent fit. After starting with the full model, the individual associations, predictive or correlational, were then examined and the associations with the least significant twotailed p values were removed from the model one by one to reach the most simplified
76 model possible that would have a CFI of .95 or greater. When the model was simplified through the stepby step reducti on process, the goodness of fit indices suggested a good fit (CFI = 1.000, TLI = 1.004, RMSEA = 0.000) and the reduced model resulted in a Chi square value of 50.935 ( p = .6669). Several predictive associations were retained in the model despite having nonsignificant p values because of their centrality to Expectancy Value theory. Specifically, many of the relationships between the following variables and high school graduation were found to be not significant, but were kept in the model nonetheless: s tudent English self efficacy scores (BYENGLSE); student control expectation scores (BYCONEXP); student academic expectation (BYSTEXP); students perceptions of teacher and school counselor expectations (BYS27H and SCE respectively); student instrumental motivation/utility value (BYINSTMO); student reading achievement score (BYTXRSTD); and student agreement with the importance of good education and how much they like school (BYS54O and BYS28 respectively). Table 43 presents the standardized coefficients for predictive associations that were retained in the model, including the nonsignificant associations that were central to Expectancy Value theory. First Research Question The first research question, how motivational factors and tenth grade reading achievement scores predict high school graduation for Hispanic males, was addressed by developing an overall model that included 10th grade Hispanic male respondents of all socioeconomic levels. As previously mentioned, the resulting reduced model for the se males across all levels of SES indicated a good model fit. This final model is depicted in Figure 42. Only the statistically significant (at the p = .05 level) predictive
77 Table 43. Predictive associations retained in the reduced model Predictive Association Estimate p value GRADUATIONYESNO on BYTXRSTD .018 .089 GRADUATIONYESNO on BYINSTMO .078 .606 GRADUATIONYESNO on BYS54O .101 .540 GRADUATIONYESNO on BYS28 .117 .351 GRADUATIONYESNO on BYENGLSE .023 .859 GRADUATIONYESNO on BYCONEXP .125 .369 GRADUATIONYESNO on BYSTEXP .023 .477 GRADUATIONYESNO on BYS27I .215 .003 GRADUATIONYESNO on BYTE20 .319 .000 BYENGLSE on BYTXRSTD .021 .000 BYENGLSE on BYINSTMO .517 .000 BYENGLSE on ENJOY .163 .000 BYCONEXP on BYTXRSTD .017 .000 BYCONEXP on BYINSTMO .691 .000 BYCONEXP on ENJOY .082 .032 BYSTEXP on BYTXRSTD .046 .000 BYSTEXP on BYINSTMO .411 .000 BYSTEXP on BYP81 .186 .000 BYSTEXP on BYS37 .455 .000 BYS27I on BYS37 .482 .000 BYS27I on BYS54O .415 .000 BYS27H on BYS54O .493 .000 BYS27H on BYS28 .376 .000 BYS27H on BYS27A .539 .000 BYS27H on BYTXRSTD .013 .006 BYS27H on ENJOY .125 .008 SCE on BYINSTMO .241 .003 Note: In MPLUS language, on is short for regressed on, and can be read predicted by. P values equal to or less than .050 are considered s tatistically significant. S ignificant associations are marked with an asterisk.
78 Figure 42. MPLUS model of predictive associations among selected ELS:2002 variables for 10th grade Latino males
79 associations are shown in this figure to simplify the depiction, even though many correlational associations in the model were also statistically significant. The nonsignif icant predictive relationships that were retained in the model (because of their centrality to Expectancy Value theory) are not depicted in Figure 42. Significant predictive r elationships High school graduation was significantly predicted by students perceptions that parents expect success in school ( p = .003) and by students English teachers educational expectations for academic attainment ( p = .000). High school graduation as predicted by students reading test standardized scores approached statistical significance ( p = .089). Many other statistically significant predictive relationships were described in the reduced model and are depicted in Figure 42 English self efficacy scores were significantly predicted by st udents instrumental motivat ion/utility interest, how much students enjoy reading and reading achievement scores Student control expectation scores were significantly predicted by reading achievement scores, instrumental motivation/utility interest score s, and how much students enjoy reading. Additionally, student attainment expectations were significantly predicted by instrumental motivation/utility interest, student reading achievement scores, the importance of good grades, as well as parents academic attainment expectations. Student agreement with the importance of good grades also significantly predicted student perceptions that parents expect success in school. Students agreement with the importance of a good education significantly predicted students perceptions that both teachers and parents expec t success in school. Furthermore, students perceptions that teachers expect success in school was
80 significantly predicted by student reading achievement scores, how much students enjoy reading, h ow much students like school, and how challenging and interesting students find classes. Student instrumental motivation/utility interest scores also significantly predicted student perceptions of school counselors academic expectations. Additional not able relationships and findings In addition to the many significant predictive relationships described, many correlational relationships were also found to be significant in the reduced model. In this reduced model, these relationships include, but are not limited to, significant correlations between residuals for English self efficacy and control expectation, English self efficacy and the perception that parents expect success in school, control expectation and the perception that parents expect success i n school, and the perception that parents expect success in school and the perception that teachers expect success in school. Furthermore, reading achievement scores were significantly correlated with instrumental motivation/utility interest, as well as w ith both parent and teacher expectations for academic attainment. Parent expectations for attainment were significantly correlated with student agreement with both the importance of good grades and the importance of a good education. Many other significant correlations were included in the reduced model, and all of the statistically significant correlations are summarized in Table 4 4. Several predictive relationships that were expected to be significant based on Expectancy Value theory did not result in statistically significant p values in the final model. For example, English self efficacy, control expectation, and instrumental motivation/utility interest did not significantly predict high school graduation for 10th grade Latino males in this study. A dditionally, parent expectations for students
81 Table 44. Statistically significant correlations in final model Correlational Association Estimate p value BYENGLSE with BYCONEXP .240 .000 BYENGLSE with BYS27I .127 .008 BYCONEXP with BYS27I .126 .011 BYS27I with BYS27H .505 .000 SCE with BYSTEXP .288 .010 BYTXRSTD with BYINSTMO 1.593 .000 BYTXRSTD with BYS27A .912 .009 BYTXRSTD with BYS54O .412 .029 BYTXRSTD with BYP81 3.256 .000 BYTXRSTD with BYTE20 7.022 .000 ENJOY with BYINSTMO .223 .000 ENJOY with BYS28 .152 .000 ENJOY with BYS27A .208 .000 ENJOY with BYS37 .145 .000 ENJOY with BYS54O .055 .002 ENJOY with BYP81 .202 .015 BYS28 with BYINSTMO .181 .000 BYS27A with BYINSTMO .180 .000 BYS27A with BYS28 .245 .000 BYS37 with BYINSTMO .305 .000 BYS37 with BYS28 .217 .000 BYS37 with BYS27A .193 .000 BYS54O with BYINSTMO .155 .000 BYS54O with BYS28 .081 .000 BYS54O with BYS27A .067 .000 BYS54O with BYS37 .173 .000 BYP81 with BYINSTMO .327 .000 BYP81 with BYS28 .185 .000 BYP81 with BYS27A .143 .004 BYP81 with BYS37 .310 .000 BYP81 with BYS54O .142 .000 BYTE20 with BYINSTMO .378 .000 BYTE20 with BYS28 .130 .002 BYTE20 with BYS27A .169 .005 BYTE20 with BYS37 .279 .000 BYTE20 with BYS54O .101 .010 BYTE20 with BYP81 .708 .000
82 educational attainment did not predict student agreement with the idea that parents expect success in school. Similarly, teacher expectations for students educational attainment did not predict student agreement with the idea that teachers expect success in school. Second Research Question The second research question addressing the extent to which these factors differ among Hispanic males from varying levels of SES was also addressed based on the overall model developed to address the first research question. Th e MPLUS computer program was also used to calculate the difference between the predictive association estimates for lower and upper SES halves of the 10th grade Latino male respondent population. Socioeconomic halves were used instead of the intended soci oeconomic quartiles to ensure that the number of cases (students) in each level of SES would be adequate for the statistical analysis. Furthermore, these halves were determined within the 10th grade Latino male population instead of in comparison to the e ntire 10th grade student population, so there was exactly half of the student sample in the lower SES half and half in the upper SES half. The predictive associations, estimates and pvalues for these associations, estimate differences and pvalues for th e differences are shown in Table 45. The differences were calculated by subtracting the higher SES predictive association estimate from the lower SES predictive association estimate, and a pvalue was obtained to determine if this difference was statisti cally significant. The only statistically significant difference between the lower and higher socioeconomic groups of 10th grade Latino males in this study was for the predictive association BYSTEXP on BYTXRSTD, or student expectation for educational attainment regressed on student reading test standardized score ( p = .001).
83 Table 45. Comparison of two tailed P values for lower and higher SES 10th grade Latino males Predictive Association Lower SES Estimate Lower SES P value Higher SES Estimate Higher SES P value Estimate Difference P value for Difference BYSTEXP on BYP81 .199 .001 .278 .001 .079 .445 GRADUATIONYESNO on BYTXRSTD .011 .468 .027 .100 .016 .464 GRADUATIONYESNO on BYINSTMO .021 .911 .276 .425 .297 .449 GRADUATIONYESNO on BYS54O .020 .926 .329 .363 .309 .461 GRADUATIONYESNO on BYS28 .098 .511 .493 .128 .395 .268 GRADUATIONYESNO on BYENGLSE .074 .682 .140 .486 .214 .429 GRADUATIONYESNO on BYCONEXP .052 .809 .224 .359 .276 .397 GRADUATIONYESNO on BYSTEXP .025 .518 .069 .182 .094 .146 GRADUATIONYESNO on BYS27I .178 .096 .329 .008 .150 .360 BYENGLSE on BYTXRSTD .024 .000 .015 .005 .009 .281 BYCONEXP on BYTXRSTD .008 .148 .021 .000 .013 .107 BYSTEXP on BYTXRSTD .063 .000 .007 .543 .055 .001 BYS27H on BYTXRSTD .021 .010 .005 .512 .017 .118 BYENGLSE on BYINSTMO .472 .000 .588 .000 .115 .294 BYCONEXP on BYINSTMO .762 .000 .655 .000 .108 .374 BYSTEXP on BYINSTMO .282 .046 .562 .001 .280 .192 SCE on BYINSTMO .266 .027 .220 .041 .045 .778 BYSTEXP on BYS37 .663 .000 .184 .276 .478 .055 BYS27I on BYS37 .391 .000 .562 .000 .171 .271 BYS27I on BYS54O .397 .008 .435 .015 .039 .867 BYS27H on BYS54O .503 .000 .454 .005 .049 .808 BYS27H on BYS28 .332 .003 .457 .000 .125 .418 BYS27H on BYS27A .547 .000 .534 .000 .013 .914 BYENGLSE on ENJOY .115 .041 .169 .008 .053 .530 BYCONEXP on ENJOY .013 .793 .131 .026 .145 .062 BYS27H on ENJOY .077 .342 .153 .013 .076 .453 GRADUATIONYESNO on BYTE20 .345 .003 .288 .024 .057 .741 Note: In MPLUS language, on is short for regressed on, and can be read predicted by. The estimate difference is calculated by s ubtracting the higher esti mate from the lower estimate. P values equal or less than .05 are considered statistically significant and are marked with an asterisk
84 In other words, the ability of student standardized reading achievement to predict student academic expectations was significantly different for lower and higher socioeconomic Latino 10th grade males. Specifically, student reading scores significantly predicted student academic expectations for lower socioeconomic 10th grade Latino males, but not for higher socioeconomic 10th grade Latino males. In other words, higher reading scores predict higher student academic expectations for lower, but not higher, SES 10th grade Latino males. Another association, BYSTEXP on BYS37, or the ability of student reported importance of good grades to predict student expectation for educational attainment approached statistical significance ( p = .055). In other words, the importance of good grades predicts student expectation for educational attainment more for lower SES 10th grade Latino males than for higher SES 10th grade Latino ma les, but this difference is not quite to a statistically significant level. All other predictive associations from the overall model were not statistically significant between lower SES and higher SES 10th grade Latino males capable of completing the ELS: 2002 questionnaire. Differences between the correlational associations for lower and higher SES were not calculated as these differences were not included in research questions guiding this study. The purpose of this study was to use Eccles et al. (1983) Expectancy Value Theory of motivation and tenth grade reading achievement scores to identify the factors that predict high school graduation for Latino males from various socioeconomic levels. In this chapter, results of the data analyses used to answer the two research questions were discussed and presented. Specifically, this study addressed the following
85 research questions: how do motivational factors and tenth grade reading achievement scores predict high school graduation for Hispanic males, and to what extent do these factors differ among Hispanic males from varying levels of SES? Descriptive statistics for key variables were presented, followed by an explanation of the factor analysis results. Additionally, the results of the structural equation modeling analysis were presented and the predictive associations were depicted graphically, with the statistically associations identified. Finally, upper and lower SES groups of 10th grade Latino males were compared and statistically significant differences were reported. In the next chapter, the results will be discussed in greater depth and applied to the work of school counselors and other educators. Additionally, directions for future research will be explored and limitations to this study will be presented.
86 CHAPTER 5 DISCUSSION The purpose of this study was to use Eccles et al. (1983) Expectancy Value Theory of motivation and tenth grade reading achievement scores to identify the factors that predict high school graduation for Latino males from various socioeconomic levels. In the previous chapter, the results of the data analyses were presented. In this chapter, these results will be discussed in depth, challenges and barriers faced by many Latino males will be presented, and implications for school counselors will be proposed. Finally, potential limitations of this study will be reviewed and directions for future research will be explored. Discussion of Results This study examined the impact of motivational variables and reading achievement on high school graduation for a nationally representative sample of 10th grade Latino males in the United States. To address this purpose, motivation as a larger construct was broken down into subconstructs based on Expectancy Value Theory (Eccles et al., 1983). In addition, reading achievement was included in the proposed model based on the extensive evidence of underachievement in reading by Latino males relative to Latina females and to students of other ethnic groups, a national trend also seen in general among low income and culturally diverse youth (Bemak & Chung, 2005; Education Trust, 2005a, 2005b; KewalRamani et al., 2007; Lopez, 2009; Planty et al., 2008). The results from this study contribute to the literature on the achievement motivations of Latino males from higher and low er socioeconomic levels, and provide insight into the complex relationships among motivation subconstructs, reading achievement, and high school graduation.
87 In this study, high school graduation was chosen as the dependent variable based on the national s tatistics that show Latino males low graduation rates relative to Latina females and students of other ethnic groups (Kewal Ramani et al., 2007; Lopez, 2009). Latino adolescents (male and female) have the highest dropout rates of any population in the U nited States with approximately 22% not completing high school, in comparison to 6% of White students and 11% of African American students (Planty et al., 2008). The data in the present study indicate that 32.2% of Latino males in the 10th grade in 2002 did not receive a high school diploma with their peers two years later. For this study, only students who graduated on time (spring/summer of 2004) or early (before spring/summer of 2004) were considered to have graduated with their peers, and this group did not include students who received a General Equivalency Diploma (GED), who dropped out, or who took longer to graduate, factors that likely contribute to the discrepancy between the high school completion rates found in the current study and other nati onal statistics. In other words, the operationalization of the graduation variable needs to be considered when one compares the results of this study to those from other studies. Predicting High School Graduation The results of this study indicated that a few key factors significantly predicted high school graduation for Latino males, two of which focus on adult expectations for student success. Specifically, students perceptions that parents expect success in school as well as English teachers educational expectations for students academic attainment significantly predicted students high school graduation status two years later. Similar findings from other studies have highlighted the connection among adult expectations, student perceptions of these expectations, and outcomes such as academic
88 achievement, especially for low income students or culturally diverse students (Benner & Mistry, 2007). The link between student perceptions of parent expectations and youths educational outcomes has been well established in the literature (Carpenter, 2008; Benner & Mistry, 2007; Davis Kean, 2005; De Civita, Pagani, Vitaro, & Tremblay, 2004; Halle, Kurtz Costes, & Mahoney, 1997). Similarly, Goldenberg et al. (2001) found that high parental academic expectations did not necessarily translate to higher student educational achievement for immigrant Latino families, and suggested that instead one must look at student perceptions of these expectations. In addition, education literature has consistently identified c onnections between teachers expectations for students success and students educational outcomes, especially for low income students and students of color (Benner & Mistry, 2007; Madon, Jussim, & Eccles, 1997). The significant impact of these adult expectations on student developmental outcomes such as educational attainment and success makes sense from an ecological systems perspective, given the interaction between person and environment (Bronfenbrenner & Morris, 1998). The present study highlights th e importance of high academic expectations from significant adults (i.e. parents, teachers, school counselors) for the educational success and attainment of Hispanic male students. Impact of Reading Achievement Reading achievement was also included in th e current study based on data indicating that male students, and culturally diverse males in particular, had lower reading achievement scores than their female or White male peers (Bemak & Chung, 2005; Education Trust, 2005a, 2005b; KewalRamani et al., 200 7; Planty et al., 2008; Lopez, 2009). Although this study found that the ability of reading achievement scores
89 to predict high school graduation for Latino males only approached statistical significance (but did not quite reach this level), this finding differs from previous studies where reading achievement was found to not predict educational attainment at all (Sciarra & Whitson, 2007; Trusty, 2002). However, reading achievement did significantly predict student English self efficacy scores, student control expectation scores, student expectation for educational attainment, as well as student perception that teachers expect success in school. These results appear to support the Expectancy Value model of student motivation by Eccles et al. (1983), in that previous experience with success (here, with reading achievement) contributes to or predicts an expectation for similar successes in the future. Additionally, student reading scores significantly predicted student academic expectations for lower socioeconomic 10th grade Latino males, but not for higher socioeconomic 10th grade Latino males. This adds to the existing literature on the educational experiences of low income students, and of low income Latino males in particular. Previous research has shown that males from low socioeconomic backgrounds may be particularly impacted by negative teacher expectations and evaluations (Auwarter & Aruguete, 2008), and that gender gaps in reading achievement are wider for low income students than for higher income students (Chatterji, 2006; Francis & Skelton, 2005). This study adds to school counselors and other educators understanding of the educational experiences of low income Latino males, highlighting the need for an increased focus on improving the reading achievement of these young men.
90 Academic Attitudes and Attainment Another finding of note from this study was that although Latino male respondents expressed p ositive attitudes about the importance of a good education, these values did not predict high school completion and were not being reflected in the graduation rates Additionally, the Latino males in this study reported generally high expectations for academic attainment, but these expectations did not predict high school graduation. These findings are similar to the findings reported by Carpenter (2008) and Goldenberg et al. (2001), where student expectations for success also were not related to achievement. This gap between knowledge of the importance of educ ational attainment versus actual hig h school graduation and college enrollment rates has been noted in other national studies. For example, recent research by the Pew Hispanic Center reflects this discrepancy: 89% of Latino students age 16 and older across the country agree strongly with the importance of higher education to succeed in l ife, but fewer than half of that group said that they actually plan to get a college degree (Lopez, 2009). However, there exists a significant gap between the college aspirations of nativeborn and foreignborn Latino students, which may partially accoun t for the previously mentioned discrepancy: 60% of nativeborn Latino students report wanting to attain at least a bachelors degree, while only 29% of foreignborn Latino students plan to do so (Lopez, 2009). Several barriers could be related to this di fference, including a need to enter the workforce to support a family as well as English language difficulties, and these challenges will be discussed more in depth in the following section. Expectancy Value Variables In this study, the impact of the Expectancy Value theory variables on graduation proved to be very complex and indirect, with more predictive relationships emerging
91 among these independent variables than anticipated. One possible explanation for this complexity is the particular question it ems used to operationalize the constructs from Expectancy Value theory, a challenge that is often encountered when using these large national databases (Sciarra & Whitson, 2007). Similarly, many of the variables in this study were composite variables created by the ELS:2002, so it makes sense that the relationships among several composite variables would be more complex than those among simpler variables. Generally, the expectancy variables and the task value variables in this study were strongly correlat ed and often predicted other independent variables more than the target dependent variable of graduation, as depicted in Figure 4 2. Interestingly, another trend that resulted from the data was that many of the task value variables predicted many of the expectancy variables for this study. Conceptually, this follows with the Expectancy Value theory in that when students place more value on a task, they may be more likely to think that adults in their lives expect success in that area, and perhaps more lik ely to expect success for themselves as well. Challenges and Barriers When considering the academic achievements and motivations of Latino male students, one must also be aware of the unique set of challenges and barriers that these students must overcome to be successful in school (Brown, Santiago, & Lopez, 2003). When students and families are aware of the challenges they face, they can become more empowered to overcome these obstacles (West Olatunji et al., 2010). Many institutional barriers and challenges have been presented in contemporary counselor education literature, some of which are faced by other marginalized populations, and some that are more specific to the Latino male population. Latino students often encounter several other barri ers along their educational path, as
92 evidenced by their underrepresentation in advanced courses (and overrepresentation in special education classes), their low high school completion rates, and their low postsecondary enrollment and completion rates (Fry, 2009; Saenz & Ponjuan, 2009; Smith Adcock et al., 2006). Issues such as high poverty levels, language barriers, and immigration status are some of the frequently cited group challenges that Latino students face (Martinez, DeGarmo, & Eddy, 2004). Latino male students face a high rate of institutional barriers that hinder their success in school. Poverty A prime example of a major systemic barrier cited in the literature and faced by many Hispanic students is poverty: Latino students in general are more l ikely than their White peers to live in poverty, and poverty presents its own set of challenges to surmount in the school setting (DotsonBlake et al., 2009; Prelow & Loukas, 2003). For example, children in poverty are more likely to attend highpoverty s chools with less experienced teachers, lower quality teaching, fewer classroom resources, larger class sizes, and higher rates of teacher attrition (Amatea & West Olatunji, 2007b; Martinez et al., 2004). Latino students are more likely to live in poverty and attend high poverty schools than White students (Prelow & Loukas, 2003), with an estimated 28% of Latino children living below the poverty line (Llagas & Snyder, 2003). Research has also indicated that many teachers have decreased academic expectations for students living in poverty, and especially for low SES males ( Auwarter & Aruguete, 2008; Diamond et al. 2004). Furthermore, other researchers have found that the intersectionality of being male, culturally diverse, and low SES predicts lower reading achievement (Entwisle et al., 2007), indicating that unique institutional barriers may impact this population. As a result, these disadvantaged educational settings often result in decreased educational
93 attainment and achievement for the large numbers o f Latino students in poverty (Amatea & West Olatunji, 2007b; Prelow & Loukas, 2003; The College Board, 2010). English Proficiency In addition to the challenges faced by students in poverty, education literature frequently indicates that having English as a second language (and more specifically, being less than proficient in English) can often serve as a barrier to educational success in the United States, as almost all schools and all standardized tests are conducted in English and demand a certain level of English proficiency, and most communication between schools and families occurs in English (DotsonBlake et al., 2009; Martinez et al., 2004; Saenz & Ponjuan, 2009; The College Board, 2010). In the current study, 43.5% of 10th grade Latino males repor ted Spanish as their native language, indicating that English may be their second language. It is important to note that simply having English as a second language does not imply a level of English proficiency. In other words, students can have a strong command of the English language even if it is their second language, and being multilingual should be viewed by school counselors and other educators as a strength rather than a weakness. However, poor English skills were reported by Latino adults over t he age of 26 as one of the most common explanations as to why Latino students do not do as well as other students in school (Lopez, 2009). Being less than proficient in English may affect student test scores, grades, and other achievement related outcomes as students may not be able to demonstrate their knowledge or understanding of a topic as well as they might in their native language, or they may not understand lessons or instructions as well as native English speakers (Eamon, 2005). Similarly, Latino adults also believe that cultural differences (including language) between Latino students and their
94 predominately White, middle class teachers represents a major reason that Latino students do not achieve as highly as their nonHispanic peers (Lopez, 2009). This view is echoed in counseling literature that speaks to the cultural discontinuity that many culturally diverse students experience in the school setting, causing them psychological distress and academic difficulties (Cholewa & West Olatunji, 2008; DotsonBlake et al., 2009). Latino students also report that not being proficient in English causes them further social and emotional stress in the school setting when other students draw attention to (or make fun of) their comprehension skills (Behnke, Piercy, & Diversi, 2004; The College Board, 2010). Finally, the Latino families in the study by Behnke et al. (2004) reported that they would appreciate and benefit from support from the school in gaining English proficiency themselves so they could better work with the school to benefit their children. While English as a second language status was not included as a variable in the model proposed in this study, it is nonetheless an important consideration for school counselors and other educators working with this population and their families given the impact that this cultural difference may have on students achievement. Other Barriers Latino families also experience several other institutional barriers that hinder academic success and attainment. Fo r example, an additional challenge faced by a significant percentage of Latino students results from the higher rate of immigrant and migrant children attending schools in the United States, as compared to their White or African American peers (DotsonBlak e et al., 2009; Fry, 2005; Gibson & Hidalgo, 2009; Saenz & Ponjuan, 2009). Latino families with migrant workers move more frequently, and may not have the consistency in educational programming that less mobile
95 students are afforded (CranstonGingras & Anderson, 1990). Additionally, undocumented immigrant children experience anxiety and fear about deportation or being identified as being an undocumented immigrant (DotsonBlake et al., 2009). Fry (2005) notes that foreignborn students are more likely to drop out of high school if they arrive to the United States later in school and if they were not making adequate academic progress in their country of origin before their move. Another barrier to academic attainment mentioned in the literature is the lack of Latino individuals (and males in particular) in the teaching workforce as role models for adolescents (Saenz & Ponjuan, 2009; The College Board, 2010). Martinez (2003) also calls for more bilingual and ethnic educators to serve as role models and as sources of information about postsecondary education for high school Latino students. Furthermore, when Latino students see school faculty and staff that represent their cultural background, their intellectual potential to succeed in high school is reinforc ed (Brown et al., 2003). Finally, families also report needing i ncreased access to information about academically related topics, including educational opportunities and options as well as financial assistance (Behnke et al., 2004; Epstein, 2007; Kelly et al., 2010; Martinez, 2003). Specifically, many Latino high school students lack adequate advising and do not get information about the necessary steps required to further their education or the many resources available to assist them in this effort (Kel ly et al., 2010; Martinez, 2003). In addition, Latino students report that teachers and counselors hold generally low academic expectations for them (Brown et al., 2003; Martinez, 2003). These low expectations may lead these educators to engage in subtle racism by not encouraging
96 these students to take higher level classes (Sciarra & Whitson, 2007) or by not sharing information about necessary pathways for advancement and postsecondary opportunities (West Olatunji et al., 2010). Furthermore, for many Latino students economic uncertainties coupled with the lack of information about educational opportunities can impact their desire or willingness to pursue a college degree (Martinez, 2003, p. 14). Finally, the aforementioned language barrier can often lead to families not feeling able or welcome to access services or information through their childrens school officials ( Sciarra & Whitson, 2007). In general, Latino students and families report experiencing more barriers to their participation and involvement in the schools than do nonLatino families (Martinez et al., 2004). Additional Challenges for Latino Males While these challenges are faced by many students within the larger Latino population, Latino male students are still underachieving in compar ison to their Latina female peers (Saenz & Ponjuan, 2009). Furthermore, Latina females have regularly reported higher academic aspirations than Latino males in the United States (Saenz & Ponjuan, 2009). The Latino gender achievement and attainment gap persists into postsecondary education: enrollment gaps in undergraduate education continue to widen between Latino males and females in the United States even though gender gaps for other ethnic groups have held constant (Fry, 2009; Fuller, 2010; Kelly et al ., 2010). Given the persisting achievement and attainment gap between Hispanic males and females, one must consider that Latino males are perhaps facing additional obstacles or pressures that contribute to their relatively low educational attainment even when compared to Latina females.
97 Discrimination Recent research in the area of the academic experiences of Latino students has pointed to the finding that Latino males report experiencing racial discrimination to a greater degree than Latina females (Alf aro, Umana Taylor, Gonzales Backen, Bamaca, & Zeiders, 2009; Lopez, 1995). Discrimination has been previously shown to negatively impact academic outcomes as well as achievement motivations for culturally diverse youth ( Alfaro et al., 2009; Eccles, Wong, & Peck, 2006 ; Wong, Eccles & Sameroff, 2003). Furthermore, discrimination in the school setting, combined with additional institutional barriers, is related to decreased academic success for Latino students specifically (Alfaro et al., 2009; DeGarmo & Mar tinez, 2006; Martinez et al. 2004). Need to support family Many researchers point to the significant pressure that young Latino males experience to start earning money as soon as possible to support their families ( Kelly et al., 2010; Lopez, 1995; Lopez 2009; Martinez et al., 2004 ; Saenz & Ponjuan, 2009). In fact, Lopez (2009) found that the biggest reason for the gap between the value placed on educational attainment and Latino male students plans to pursue postsecondary education is the pressure to support a family financially. The Pew Hispanic S tudy emphasizes that the expectation within Latino cultures that Latino males enter the work force to support their families as soon as possible exists significantly earlier than within other ethnic and gend er groups (Lopez, 2009). Additionally, the desire and pressure to fulfill the traditional gender role of provider is often even stronger for foreignborn Latino male adolescents as compared to nativeborn Latino males (Fry, 2005). Furthermore, the tradit ional Hispanic value of familismo is often salient in the minds of young Latino males, a value that encourages a strong attachment, loyalty, and responsibility to
98 immediate and extended family ( Kelly et al., 2010; Saenz & Ponjuan, 2009 ; The College Board, 2010). This particular value of fidelity to the family could conflict with the value Latino males place on pursuing postsecondary education, and personal sacrifices by these young men are often made to promote the well being of the family ( Kelly et al., 2 010; Saenz & Ponjuan, 2009). Ultimately, the social, familial, and socioeconomic pressures faced by young Latino males (foreign born or native born) may manifest themselves in the decision to join the workforce earlier than their Latina female peers, in definitely passing up the opportunity to seek a postsecondary education (Saenz & Ponjuan, 2009, p. 63). Implications for School Counselors The current study adds to previous research on the educational attainment of Latino males, and the results from this study and others can be used to inform the work that school counselors (and other educators) do with these young men. In general, the ASCA model encourages school counselors to increase their awareness of cultural differences and become advocates for s tudents, in particular for students who are marginalized in the school system and in society, and to work to eliminate the barriers to success for these students (Amatea & West Olatunji, 2007b; Bemak, 2000; Bemak & Chung, 2005; Lee, 2001; West Olatunji et al., 2010). School counselors are encouraged to promote success for all students by integrating social, psychological, career, and academic development, and high school graduation represents a key benchmark of this success (ASCA, 2005). By examining quantitative data that demonstrate achievement and attainment patterns of Latino students, school counselors in a leadership position can better understand these students educational experiences and can collaborate with other stakeholders to make necessary e vidence-
99 based improvements (Amatea & West Olatunji, 2007a; Bemak, 2000; Bryan, 2005; Clark & Stone, 2007; Lee, 2001; SmithAdcock et al., 2006; Stone & Dahir, 2006). In these ways, school counselor efforts can be directly connected to educational outcomes such as high school graduation. Translating theoretical implications into practical recommendations is essential for school counselors working with Latino males and their families to eliminate the existing systemic barriers and ensure the academic success of not only this population, but also of all students ( Goh et al., 2007; Rodriguez & Morrobel, 2004). Creating a culturally responsive, asset based developmental school counseling and guidance program involves school counselors acting in leadership posit ions, serving as student advocates, and connecting schools with the families and communities they serve (Amatea & West Olatunji, 2007b; Bemak & Chung, 2005; Lee, 2001; Stone & Dahir, 2006). With Latino students now representing the largest and fastest growing culturally diverse student population (Bernstein, 2008; Planty et al., 2008; Sciarra & Whitson, 2007; U.S. Census Bureau, 2008), these skills are especially necessary for school counselors to possess and implement (SmithAdcock et al., 2006). School Counselors as Educational Leaders Within the school walls, school counselors have an obligation to operate in a leadership position and serve as student advocates to help all students succeed (ASCA, 2005). To this end, culturally responsive school counsel ing services (based in the worldviews of the students and families served) are necessary to provide equal educational access and opportunities, especially to those student populations who are not experiencing the same academic success and attainment as mor e advantaged students ( Goh et al., 2007; Lee, 2001; Smith Adcock et al., 2006). Essentially, the goal
100 of school counselors culturally responsive and asset based interventions for the Latino population is to foster developmentally appropriate environment s that embrace the culturally unique strengths of Latino youths in ways to enhance their ability to take advantage of the assets they have (Rodriguez & Morrobel, 2004, 121). Counseling research and literature reveals many effective means to provide culturally responsive school counseling services, including working with students, teachers, and administrators to promote academic success and improve high school graduation for all students. For example, through their understanding of systemic contexts and power relationships, school counselors as leaders and change agents can assist students directly in understanding the challenges they face, thereby empowering them to overcome these barriers (West Olatunji et al., 2010). Additionally, in their positio ns as leaders in the academic setting, school counselors have a responsibility to be aware of their own influences on these students, as they may unintentionally perpetuate the previously discussed institutional barriers and limit the success of Latino mal e students or other socially marginalized students (Bemak & Chung, 2005; West Olatunji et al., 2010). In this study, students perceptions of their school counselors expectations for them did not predict graduation, so perhaps school counselors are not a dequately or effectively communicating their own high expectations for Latino males. An increased awareness of their own influences and expectations can lead to school counselors encouraging other educators and parents to set (and vocalize) high academic expectations for all students, and for Latino males in particular. All educators must actively confront the frequently reported low expectations for Latino students to help establish a rigorous academic program for postsecondary preparation (Brown et
101 al., 2003; Kelly et al., 2010), and the current study indicates that Latino male high school students can particularly benefit from higher expectations by teachers and other significant adults in their lives. School counselors have unique training in interper sonal communication, and can work with staff to improve both the teacher student relationship as well as the teachers communication of academic expectations. School counselors as leaders and advocates in the academic setting can also work to educate teachers about the institutional barriers present in school settings that disadvantage Latino students (Brown et al., 2003) to offer insight or alternative perspectives to teachers about the academic behaviors of culturally diverse or low SES students and their families, to assist teachers with identifying student/family/community assets that can be maximized in the classroom, and to aid all educational staff in efforts to effectively communicate and collaborate with families (Amatea & West Olatunji, 2007b; Goh et al., 2007). School counselors can also collaborate with teachers to develop culturally responsive and relevant curriculum and lessons (Amatea & West Olatunji, 2007b; Clark & Stone, 2007; Goh et al., 2007). Many authors have also suggested that s cho ol counselors can promote the achievement and motivation of culturally diverse students by implementing culturally responsive programs that connect students with community members from diverse backgrounds, including programs such as job shadowing, mentoring, or internships with business partners, community leaders, and other similar role models (Bryan, 2005; Clark & Stone, 2007; Lee, 2001; The College Board, 2010). Additionally, Martinez (2003) reports that successful programs for improving the academic ex pectations, plans and motivations of Latino students also include aspects such as mentoring, tutoring, peer advising, and college preparatory
102 programs programs that can be initiated by proactive school counselors. Finally, school counselors need to acti vely disseminate information about programs that have been effective in improving the academic motivation and achievement in their schools through professional journals and meetings (Brown et al., 2003). School Family Partnerships Additionally, school co unselors in a leadership role can and should coordinate efforts to connect schools with the families they serve in meaningful ways (Clark & Stone, 2007; Epstein, 2007; Goh et al., 2007; Stanard, 2003). The current study indicates that the perception by Latino male students of high parental academic expectations significantly predicts high school graduation. Thus, counselors can work with families of Latino students to promote high expectations, access to information, and opportunities for involvement. Sp ecifically, school counselors can facilitate effective communication between parents and students so students are able to articulate that their parents expect a certain level of educational attainment from them. In this role, school counselors can employ their understanding of communication and relationships to potentially enhance Latino males academic motivation and success. In addition whenever possible or necessary, using a translator can assist nonSpanish speaking school counselors to better connect and communicate with Latino students, families, and community members, bridging the language gap that often distances these individuals (Sciarra & Whitson, 2007). School counselors can also make written materials available in Spanish as well as English and encourage fellow educators to do the same, especially when sending home written communication pertaining to academic needs and opportunities (SmithAdcock et al., 2006). School administrators also report a need for more Spanish speaking school counselors to help
103 reduce the cultural barriers for Latino students and families and to best serve the academic, social, and career needs of these children (SmithAdcock et al., 2006). An additional way that school counselors can help to bridge the cultural and language gap between Latino families and schools is to develop partnerships with paraprofessionals (Lee, 2001) or cultural brokers (DotsonBlake et al., 2009) who are familiar with the language and customs of the families and communities served by the s chools, and actively involve these individuals in family school counseling initiatives (Goh et al., 2007) Furthermore, it has been suggested that school counselors can shift the hours in their work day to allow for after school day visits to families who may have difficulty coming to the school during t ypical school hours (Lee, 2001; Sciarra & Whitson, 2007 ; Smith Adcock et al., 2006). Once again highlighting the importance of high academic expectations by significant adults in Latino students lives, r esults from a study by Behnke et al. (2004) indicate that when Latino parents have higher academic and occupational aspirations for themselves, these aspirations may transfer to, or influence, the aspirations of their children. Given this insight, educati onal programs that work with multiple generations of Latino families may positively impact the aspirations and motivations of Latino young men (Behnke et al., 2004). Furthermore, many Latino families reported simply not knowing about the educational pathw ays needed to achieve desired occupational outcomes ( Behnke et al. 2004) P rograms that effectively provide Latino families with specific information about requirements for high school completion, postsecondary opportunities and applications, and access to financial aid are imperative and would be
104 extremely valuable in improving postsecondary educational enrollment and completion ( Behnke et al. 2004; Kelly et al., 2010). Specific ways counselors can provide increased access to information include offering guidance services in community centers or places of worship in the neighborhoods where students li ve and socialize (Lee, 2001). Providing this information by going to the families and communities instead of forcing them to come to the school could help to bridge the gap between Latino families and the schools and help to reduce anxiety or fears about interacting with school officials (Lee, 2001). This effort by school counselors indicates that they are sensitive to the possible distrust of the school system that some families may have, demonstrates that the school (as represented by the school counsel or) is willing to engage families in a variety of locations and cultural contexts, and shows supportive of family involvement in students education (Lee, 2001). Providing information to and involving both Latino male students and families in culturally r esponsive ways can improve communication between home and school and may lead to higher parental expectations for success, which appear to be strong motivators for student achievement. Another consideration for school counselors when working with the Lat ino population is the high rate of migrant families and students that are present in certain areas of the country. These families have unique needs, and many school counselors may not know how to work effectively with these students, or may not even be aw are of the migrant status of their students (CranstonGingras & Anderson, 1990). To best serve and advocate for these students, school counselors need to make themselves aware of the various federal and local educational assistance programs available in
105 t heir area for these migrant families (CranstonGingras & Anderson, 1990; Gibson & Hidalgo, 2009). Additionally, school counselors can conduct a small group for migrant students to allow a space for these students to create connections to and a sense of co mmunity with other students in potentially similar life situations (CranstonGringas & Anderson, 1990). School counselors can also connect migrant students with successful adult role models who may have also been children of migrant families when they wer e younger (Gibson & Hidalgo, 2009). Finally, after getting to know the migrant communities around them, school counselors in areas with a high migrant student population can conduct inservice informational programs to educate teachers about the needs and assets of this population, thereby advocating for these students (CranstonGringras & Anderson, 1990). Limitations of the Study Because of the focus on high school Latino males in the United States, one limitation of this study is the inability to gener alize any findings to females, or to males from other ethnic groups, countries, or educational systems. These findings may not be applicable to any Latino males who are not in high school or who are not part of the national database being examined. Anoth er limitation of this study is the reliance upon one data source and the subsequent assumption that the data from this source are accurate representations of students motivation to achieve. Furthermore, using a national survey like the ELS:2002 does not allow a researcher to capture the contextual meanings, or qualitative aspects, of the real life experiences of the individuals involved (Lopez, 2003). These national surveys also tend to treat race and gender as a priori rather than ongoing processes of s ocialization and development (Lopez, 2003). Finally, constructs like motivation, as well as the subconstructs from Expectancy Value theory
106 like expectancies and task values, are relatively complex in nature, and are difficult to capture using multiple choice questionnaire items from national databases designed to measure a range of attitudinal and behavioral attributes. As with many quantitative studies that use data from a large, nationally representative database, caution must be taken when drawing conclusions about groups of people based on the results of the statistical analyses. Specifically, researchers must avoid the tendency to assume that each member of a group embodies the characteristics that might describe the group as a whole (Tinto, 1987). For example, one cannot assume that the general achievement or motivation patterns of a particular subpopulation apply to a specific student who identifies as a member of that group. However, data on the patterns of achievement motivation can provide an understanding of general group differences while also offering a springboard for discussions about the complex interaction of factors predicting graduation for groups of high school males. Finally, group data can be useful for the development of theory and policy recommendations (Tinto, 1987), and to understand the general experience of a subpopulation in order to translate theory into practice. Directions for Future Research While this study offers contributions to the existing literature on the achiev ement of Latino males, more research still needs to be conducted to further understand the educational experiences of this diverse group of students. In general, future research should aim to better inform school counselors as to how they can use their leadership positions to promote multicultural awareness and competence, as well as to work toward the elimination of barriers to educational success for all students. Additionally, future studies should recognize that Latino males, as a group, are not monol ithic, and
107 that great variety exists in the cultural backgrounds, experiences, and perspectives across the various Latino populations (The College Board, 2010) Research that acknowledges the more specific cultural backgrounds, immigrant status, and perspectives of a particular subpopulation would contribute to a more detailed understanding of the educational motivations and experiences of these specific groups (Carpenter, 2008; Eamon, 2005; Lopez, 2003). Additionally, researchers need to include and report findings for Latino participants in their studies to broaden and deepen counselors and other educators understanding of this growing population (Rodriguez & Morrobel, 2004; Brown et al., 2003). Furthermore, much of the existing research focusing on Latino populations uses a deficit oriented approach, whereas future research, in order to contribute to the development of theory, needs to continue to originate from a strengths based theoretical perspective that considers context as well as more specific cultural backgrounds (Amatea et al., 2006; Bryan, 2005; Galassi & Akos, 2007; Rodriguez & Morrobel, 2004; Walsh et al., 2007). Research in the area of Latino educational experiences and motivations could greatly benefit from more qualitative studies that could provide a more rich description of the lived experiences of this large and diverse student population. Studies like the ones by Lopez (2003) and Behnke et al. (2004) offer insight into the qualitative aspects of student and family educational experiences that are difficult, if not impossible, to understand purely through quantitative research. For example, students own comments can shed light on their perceptions of strengths they possess, what motivates them, and what barriers they experience in their school lives. Additionally, hardto -
108 define constructs like masculinity and family expectations that may impact Latino male achievement motivations are extremely difficult concepts to understand through quantitative methods, and more detail can be g ained through openended questions and interviews. Finally, there is relatively little research that focuses specifically on the Latino male experience in either high school or postsecondary education, and qualitative research could help to fill this gap (Rodriguez & Morrobel, 2004; Saenz & Ponjuan, 2009). Additional research, both quantitative and qualitative, can continue to use an ecological systems perspective to focus on the complex interactions of influences on the educational outcomes of Latino male students (Benner & Mistry, 2007). Results from the present study clearly point to an intricate web of both personal characteristics and contextual or systemic factors and experiences that predict academic motivation and success. Future research that r ecognizes the direct and indirect influences on development could help to better clarify these processes and interactions, which could then aid school counselors and other educators in their work with these students and families (Amatea & West Olatunji, 2007a). Also research in the area of school counselors roles as leaders and advocates in the school system, as well as their impact on the achievement of culturally diverse and low income students, is imperative. As previously stated, a significant body of research has pointed out the effect of adult expectations on student achievement and attainment, particularly for low income and culturally diverse student populations (Benner & Mistry, 2007; Carpenter, 2008; De Civita et al., 2004; Davis Kean, 2005; H alle et al., 1997 ; Lopez, 1995). For example, positionality research has the potential to draw attention to
109 the ways in which school counselors (and other educators) impact the achievement motivations and experiences of Latino males and other traditionall y disenfranchised students (West Olatunji et al., 2010). Finally further research using the ELS:2002 database can build on the results of this study. For exampl e, another study could compare motivation variables between the 10th grade and 12th grade years to see how these aspects of motivation change over the two years as students progress through high school and approach graduation. Further studies can also more closely examine aspects of relationships between students and the ir parents, teachers, and school counselors to better unders tand how students perceive adults expectations for their future. More data from future studies can expand the understanding of the motivational patterns of Latino male students, and can inform school counselors evidencebased practice within the school setting. Clearly, many avenues of research can continue to build o n the findings of this study and others, and have the potential to significantly improve the educational experiences and prospects of Latino male high school students. Conclusion The previous chapters have described this research study in detail. The purpose of this study was to use Eccles et al. (1983) Expectancy Value Theory of motivation and tenth grade reading achievement scores to identify the factors that predict high school graduation for Latino males from various socioeconomic levels. Chapter 1 provided an introduction to the issue of Latino male underachievement as measured by significantly lower high school graduation rates than any other demographic population in the United States. Chapter 2 presented a literature review in which Latino male achievement was explored and Expectancy Value Theory was explained in detail (Eccles et al., 1983).
110 Chapter 3 presented information about the ELS:2002 data to be used in this study and the data analysis procedures to be conducted, and Chapter 4 detailed t he results of these analyses. Finally, in this chapter, the results from the data analyses were discussed in depth, and implications for school counselors and other educators were explored. Additionally, limitations of this study were explained and direc tions for future research with Latino males were proposed.
111 REFERENCES Adebayo, B. (2008). Gender gaps in college enrollment and degree attainment: An exploratory analysis. College Student Journal 42(1), 232237. Alfaro E. C., Umana Taylor, A. J., Gonzales Backen, M. A., Bamaca, M. Y., & Zeiders, K. H. (2009). Latino adolescents academic success: The role of discrimination, academic motivation, and gender. Journal of Adolescence, 32(4), 941 962. Amatea, E., & Sherrard, P. (1994). The ecosystem ic view: A choice of lenses. Journal of Mental Health Counseling, 16 (1), 6 21. Amatea, E., Smith Adcock, S., & Villares, E. (2006). From Family Deficit to Family Strength: Viewing Families' Contributions to Children's Learning from a Family Resilience Per spective. Professional School Counseling, 9 (3), 177189. Amatea, E., & West Olatunji, C. (2007a). Rethinking how school counselors work with families and schools: An ecosystemic approach. In J. Wittmer & M. Clark. (Eds.), Managing your school counseling program: K 12 Developmental strategies (3rd ed., pp. 211222), Minn eapolis, MN: Educational Media. Amatea, E., & West Olatunji, C. (2007b). Joining the Conversation about Educating Our Poorest Children: Emerging Leadership Roles for School Counselors in Hi gh Poverty Schools. Professional School Counseling, 11 (2), 81 89. American School Counselor Association. (2005). The ASCA national model: A framework for school counseling programs. Alexandria, VA: Author. Aron, A. & Aron, E. (1999). Statistics for Psychology (2nd ed.). Sa ddle River, NJ: PrenticeHall. Atkinson, J. W. (1964). An Introduction to Motivation. Princeton, NJ: Van Nostrand. Auwarter, A E. & Aruguete, M. S. (2008). Effects of student gender and socioeconomic status on teac her perceptions. Journal of Educational Research, 101 (4), 243246. Bailey, D. F. & Paisley, P. O. (2004). Developing and nurturing excellence in African American male adolescents. Journal of Counseling and Development, 82(1), 10 17. Ba nks, M. & Woolfson, L. (2008). Why do students thi nk they fail? The relationship between attributions and academic self perceptions. British Journal of Special Education, 35(1), 49 56.
112 Behnke, A. O., Piercy, K. W. & Diversi, M. (2004) Educational and occupational aspirations of L atino youth and their parents. Hispanic Journal of Behavioral Sciences, 26(1), 16 35. Bemak, F. (2000). Transforming the role of the counselor to provide leadership in educational reform through collaboration. Professional School Counseling, 3 (5 ) 323331. Bemak, F., & Chung, R. C. (2005). Advocacy as a critical role for urban school counselors: Working toward equity and social justice. Professional School Counseling, 8 (3) 196 202. Bernstein, R. (2008). U.S. Hispanic population surpasses 45 million: now 15 percent of total. Available from http://www.census.gov/Press Release/www/releases/archives/population/011910.html Blackhurst, A., & Auger, R. ( 2008). Precursors to the gender gap in college enrollment: Children's aspirations and expectations for their futures. Professional School Counseling, 11 (3), 149 158. Blair, S. L., Blair, M. C. L., & Madamba, A. B. (1999). Racial/e thnic differences in high school students' academic performance: Understanding the interweave of social class and et hnicity in the family context. Journal of Comparative Family Studies, 30(3), 539 555. Bronfenbrenner, U. (1986). Ecology of the family as a context for human development: Research perspectives. Developmental Psychology, 22(6), 723742. Bronfenbrenner, U., & Morris, P. (1998). The ecology of developmental processes. In W. Damon & R. M. Lerner (Eds.), Handbook of child psychology (pp. 993 1028). New York: Wiley. Bryan, J. (2005). Fostering educational resilience and achievement in urban schools through school family community partnerships. Professional School Counseling, 8 (3), 219 227. Carey, J. C. & Martin, I. (2007). What are the implications of possible selves research for school counseling practice? School Counseling Research Brief 5.2. Amherst, MA: Center for School Counseling Outcome Research. Carpenter II, D. M. (2008). Expectations, aspirations, and achievement among Latino students of immigrant families. Marriage & Family Review, 43 (1 2), 164185. Carpenter II, D., Ramirez, A., & Severn, L. (2006). Gap or gaps: Challenging the singular definition of the achievement gap. Education & Urban Society, 39(1), 113127.
113 Center on Education Policy. (2008). State High School Exit Exams: A Move Toward End of Course Exams. Retrieved March 20, 2010 at http://www.cep dc.org/document/docWindow.cfm?fuseaction=document.viewDocument&docume ntid=244&documentFormatId=4224. Chatterji, M. (2006). Reading achievement gaps, correlates, and moderators of early reading achievement: Evidence from the Early Childhood Longitudinal Study (ECLS) kindergarten to first grade sample. Journal of Educational Psychology, 98(3), 489 507. Childrens Defense Fund. (2005). The state of Americas children. Retrieved October 3, 2008 at http://www.childrensdefense.org/childresearchdatapublications/data/state of americas children 2005report.html Cholewa, B. & West Olatunji, C. (2008). Exploring the relationship among cultural discontinuity, psychological distress, and academic outcomes with low income, culturally diverse students. Professional School Counseling, 12(1), 54 61. Clark, M. A., Flower, K., Wa lton, J., & Oakley, E. (2008). Tackling male underachievement : Enhancing a strengths based learning environment for middle school boys. Professional School Counseling, 12(2), 127 132. Clark, M. A., Oakley, E., & Adams, H. (2006). The gender achievement gap challenge. ASCA School Counselor, 43(3), 20 25. Clark, M. A. & Stone, C. (2007). The Developmental School Counselor as Educational Leader. In J. Wittmer & M. Clark. (Eds.), Managing your school counseling program: K 12 Developmental strategies (3rd ed., pp. 8289), Minn eapolis, MN: Educational Media. Cokley, K. O. (2003). What do we know about the motivation of African American students? Challenging the anti intellectual myth. Harvard Educational Review, 73(4), 524 558. Cooper, H., Nye, B., Charlton, K., Lindsay, J., & Greathouse, S. (1996). The effects of su mmer vacation on achievement test scores: a narrative and metaanalytic review. Review of Educational Research, 66(3), 227 268. Darling Hammond, L. (2000). New standards and old inequalities: School reform and the education of African American students. T he Journal of Negro Education, 69(4), 263 287. Davis Kean, P. E. (2005). The influence of parent education and family income on child achievement: The indirect role of parental expectations and the home environment. Journal of Family Psychology, 19 (2), 29 4 304.
114 De Civita, M., Pagani, L., Vitaro, F., & Tremblay, R. E. (2004). The role of maternal educational aspirations in mediating the risk of income source on academic failure in children from persistently poor families. Children and Youth Services Revie w, 26 (8), 749 769. DeGarmo, D. S. & Martinez Jr., C. R. (2006). A cultur ally informed model of academic well being for Latino youth: The importance of discriminatory experiences and social support. Family Relations, 55(3), 267 278. Diamond, J. B., Randolph, A., & Spillane, J. P. (2004). Teachers expectations and sense of responsibility for student learning: The importance of race, class, and organizational habitus. Anthropology and Education Quarterly, 35(1), 75 98. Dimmitt, C. (2003). Transforming school counseling practice through collaboration and the use of data: A study of ac ademic failure in high school. Professional School Counseling, 6(5), 340350. Dollarhide, C. (2003). School counselors as progr am leaders: Applying leadership contexts to school counseling. Professional School Counseling, 6 (5), 304 308. DotsonBlake, K., Foster, V., & Gressard, C. (2009). Ending the silence of the Mexican immigrant voice in public education: Creating culturally inclusive family school community partnerships. Professional School Counseling, 12 (3), 230 239. Duncan, G. J., & Brooks Gunn, J. (2001). Poverty, welfare reform, and childrens achievement. In B. J. Biddle ( Ed.), Social Class, Poverty, and Achievement (pp. 4975). New York: RoutlageFarmer. Eamon, M. K. (2005). Social demographic, school, neighborhood, and parenting influences on the academic achievement of Latino young adolescents. Journal of Youth and Adol escence, 34(2), 163 174. Eccles, J., Adler, T. F., Futterman, R., Goff, S. B., Kaczala, C. M., Meece, J. L., & Midgley, C. (1983). Expectancies, values, and academic behaviors. In J. T. Spence (Ed.), Achievement and Achievement Motives: Psychological and Sociological Approaches (pp. 75146). San Fransisco: Freeman. Eccles, J. S., Wong, C. A., & Peck, S. C. (2006). Ethnicity as soc ial context for the development of African American adolescents. Journal of School Psychology, 44(5), 407 426. Education Trust. (2005a). Stalled in secondary: A look at student achievement since the No Child Left Behind Act. Retrieved April 8, 2009 at http://www.edtrust.org/dc/publication/stalled in secondary 0
115 Education Trust. (2005b). The funding gap 2005: Low income and minority students shortchanged by most states. Retrieved April 8, 2009 at http://www.edtrust.org/d c/publication/the fundinggap report 2005. Entwisle, D. R., Alexander, K. L., & Olson, L. S. (2007) Early schooling: The handicap of being poor and male. Sociology of Education, 80(2), 114 138. Epstein, J. (2007). Connections count: Improving family and community involvement in secondary schools. Principal Leadership, 8(2), 1622. Florida Department of Education. (2007). Floridas Guide to Public High School Graduat ion. Retrieved August 28, 2009 at http://www.fldoe.org/APlusPlus/pdf/MAJORSGuideHSGraduation2007.pdf Flowers, T. A. (2003). Exploring the influence of reading for pleasure on African American high school students reading achievement. High School Journal, 87( 1), 58 62. Flowers, T. A. & Flowers, L. A. (2008). Factors affect ing urban African American high school stu dents achievement in reading. Urban Education, 43(2), 154 171. Flowers, L. A., Milner, H. R., & Moore III, J. L. (2003). Effects of locus of contr ol on African American high school seniors' educational aspirations: Implications for preservice and inservice high s chool teachers and counselors. High School Journal, 87(1), 39 50. Francis, B. (2000). Boys, girls, and achievement: Addressing the classroom issues Lon don; New York: RoutledgeFalmer. Francis, B., & Skelton, C. (2005). Reassessing gender and achievement: Questioning contemporary key debates London; New York: Routledge. Freeman, C. E. (2004). Trends in educational equity of girls and women: 2004 (NCES 2005016). U. S. Department of Education, National Center for Educational S tatistics. Washington, D.C.: U.S. Government Printing Office. Fry, R. (2005). The Higher Dropout Rate of ForeignBorn Teens: The Role of Schooling Abroad. Washingt on, D C: The Pew Hispanic Center Retrieved April 14, 2010 at http://pewhispanic.org/reports/report.php?ReportID=55. Fry, R. (2009). The Changing Pathways of Hispanic Youth into Adulthood. Washington, D C: The Pew Hispanic Center Retrieved April 14, 2010 at http://pewhispanic.org/reports/report.php?ReportID=114
116 Fuller, A. (2010). Female undergraduates continue to outnum ber men, but gap holds steady. Chronicle of Higher Education. Retrieved June 15, 2010 at http://chronicle.com.lp.hscl.ufl.edu/article/FemaleUndergraduates Conti/63726/ Galassi, J.P. & Akos, P. (2007). Strengths based school counseling: Promoting student development and achievement. Malwah, NJ: Lawrence Erlbaum Associates. Gibson, M. A. & Hidalgo, N. (2009). Bridges to success in high school for migrant youth. Teachers College Record, 111(3), 683711. Ginsberg, A. E., Brown, S. P., & Shapiro, J. P. (2004). Gender in urban education: Strategies for student achievement Portsmouth, NH: Heinemann. Goh, M., Wahl, K. H., McDonald, J. K., Brissett, A. A., & Yoon, E. (2007). Working with immigrant students in schools: The role of school counselors in building cross cultural bridges. Journal of Multicultural Counseling and Development, 35(2), 66 79. Goldenberg, C., Gallimore, R., Reese, L., & Garnier H. (2001). Cause or effect? A longitudinal study of immigrant Latino parents aspirations and expectations, and their childrens school performance. American Educational Research Journal, 38(3), 547 582. Gorard, S., & Smith, E. (2004). What is 'underachievem ent' at school? School Leadership & Management, 24(2), 205 225. Gordan, K. A. (1995). Self concept and motivational patterns of resilient African American high school students. Journal of Black Psychology, 21(3), 239255. Gordon Rouse, K. A. & Austin, J. T. (2002). The relationship of gender and academic performance to motivation: W ithin ethnic group variations. The Urban Review, 34(4), 293 316. Gurian, M. & Stevens, K. (2005). The minds of boys: Saving our sons from falling behind in school and life. San Francisco: Jossey Bass. Gutman, L. M., Sameroff, A. J., & Cole, R. (2003). Aca demic growth curve trajectories from 1st grade to 12th grade: Effects of multiple social risk factors and preschool child factors. Developmental Psychology, 39 (4), 777790. Halle, T. G., Kurtz Costes, B., & Mahoney, J. L. (1997). Family influences on school achievement in low income African American children. Journal of Educational Psychology, 89 (3), 527537.
117 HampdenThompson, G., Warken tien, S., & Daniel, B. (2009). Course credit accrual and dropping out of high school, by student ch aracteristics (NCES 2009 035). Statistics in Brief. National Ce nter for Education Statistics. Washington, D C: U.S. Department of Education. Haycock, K. (2006). Promise abandoned: How policy choices and institutional practices restrict college opportunities. Retrieved April 8, 2009 at www2.edtrust.org/NR/rdonlyres/B6772F1A 116D 4827A326 F8CFAD33975A/0/PromiseAbandonedHigherEd.pdf Washington, D. C.: Education Trus t. Howard, T. C. (2003). "A Tug of War for Our Minds:" African American high school students' perceptions of their academic identities and college aspirations. High School Journal, 87(1), 4 17. H urtado, A. & Sinha, M. (2008). More than men: Latino femini st mascul inities and intersectionality. Sex Roles, 59(5 6), 337349. Ingels, S. J., Pratt, D. J., Rogers, J., Siegel, P. H., & Stutts, E. S. (2004). Education Longitudinal Study of 2002: Baseyear data file users manual (NCES 2004 405). U. S. Department of Education, National Center for Education Statistics. Washington, D. C.: U. S. Government Printing Office. Ingels, S.J., Pratt, D.J., Wilson, D., Burns, L.J., Currivan, D., Rogers, J.E., and Hubbard Bednasz, S. (2007). Education Longitudinal Study of 200 2: BaseYear to Second Follow up Data File Documentation (NCES 2008 347). U.S. Department of Education. Washington, D C: National Center for Education Statistics. Ingels S. J. & Scott, L. A. (2004). The high school sophomore class of 2002: A demographic description first results from the base year of the Education Longitudinal Study of 2002 (NCES 2004 371). U.S. Department of Education, National Ce nter for Education Statistics. Washington, D.C.: U. S. Government Printing Office. Jianzhong, X. (2006). Gend er and homework management reported by high school students. Educational Psychology, 26(1), 73 91. Kelly, A. P., Schneider, M., & Carey, K. (2010). Rising to the Challenge: Hispanic College Graduation Rates as a National Priority. Washington, D.C.: Ameri can Enterprise Institute. KewalRamani, A., Gilbertson, L., Fox, M., & Provasnik, S. (2007). Status and Trends in the Education of Racial and Ethnic Minorities (NCES 2007 039). National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education. Washington, D .C.
118 Kline, R. B. (2005). Principles and Practice of Structural Equation Modeling, 2nd ed. New York: Guilford Press. Kobrin, J. L., Patterson, B. F., Shaw, E. J., Mattern, K. D. & Barbuti, S. M. (2008). Validity of the SAT for predicting first year college grade point average. College Board Research Report 20085. Retrieved September 14, 2009 at http://profe ssionals.collegeboard.com/datareports research/cb/validity of sat predicting fycgpa Lee, C. C. (2001). Culturally responsive school counselors and programs: Addressing the needs of all students. Professional School Counseling, 4(4), 257 261. Lee, C. C (2005). Urban school counseling: Context characteristics, and competencies. Professional School Counseling, 8 (3), 184 188. Llagas, C., & Snyder, T. D. (2003). Status and trends in the education of Hispanics. Washington, D C: U.S. Department of Education, National Center for Educational Statistics. Lopez, E. M. (1995). Challenges and resources of MexicanAmerican students within the family, peer group, and univer sity: Age and gender patterns. Hispanic Journal of Behavioral Sciences, 17(4), 499 508. Lopez, M. N. (2009). Latinos and education: Explaining the attainment gap. The Pew Hispanic Center. Retrieved April 14, 2010 at http://pewhispanic.org/reports/report.php?ReportID=115 Lopez, N. (2003). Hopeful Girls, Troubled Boys: Race and Gender Disparity in Urban Education. New York: Taylor & Francis. Ma, X. (2000). Socioeconomic gaps in academic achievement within schools: Are they co nsistent across subject areas? Educational Research and Evaluation, 6(4), 337355. MacMillan, D. L., Gresham, F. M., Lopez, M. F., & Bocian, K. M. (1996). Comparison of students nominated for prereferral interventions by ethnicity and gender. The Journal of Special Education, 30(2), 133 151. Madon, S., Jussim, L., & Eccles, J. (1997). In search of the powerful self fulfilling prophecy. Journal of Personality & Social Psychology, 72(4), 791809. Mandel, H. P., Mar cus, S. I., & Dean, L. (1995). Could do better: Why children underachieve and what to do about it. New York: John Wiley & Sons.
119 Marsh, H. W., & Yeung, A. S. (1998). Longitudinal structural equation models of academic self concept and achievement: Gender differences in the development of math and English constructs. American Educational Research Journal, 35(4), 705739. Martinez, M. D. (2003). Missing in action: Reconstructing hope and possibility among Latino students placed at risk. Journal of Latinos and Education, 2(1), 1321. Martinez, C. R., DeGarmo, D. S., & Eddy, J. M. (2004) Promoting academic success among Latino youths. Hispanic Journal of Behavioral Sciences, 26(2), 128151. McIntyre, T. & Tong, V. (1998). Where the boys are: Do cross gender misunderstandings of language use and behavior patterns contribute to the overre presentation of males in programs for students with emotional and behavioral disorders? Education & Treatment of Children, 21(3), 321 333. McCall, R. B. (1994). Academic underachievers. Current Directions in Psychological Science, 3(1), 15 19. Meece, J. L., Glie nke, B. B., & Burg, S. (2006). G ender and motivation. Journal of School Psychology, 44 (5), 351373. Meece, J. L. & Kurtz Costes, B. (2001). Introduction: The schooling of ethni c minority children and youth. Educational Psychologist, 36(1), 1 7. Moss, G. (2000). Raising boys attainment in reading: Some principles for intervention. Reading, 34(3), 101 106. National Center for Educational Statistics (2002). The condition of education. Retrieved October 9, 2005 at http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2002025. National Center for Educational Statistics (2005). The condition of education. Retrieved November 11, 2005 at http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2005094. National Center for Educational Statistics (2006). The condition of education. Retrieved September 27, 2008 at http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2006071. National Center for Educational Statistics (2007). The condition of education. Retrieved September 27, 2008 at nces.ed.gov /pubsearch/pubsinfo.asp?pubid=2007064. Neal, L. I., McCray, A. D., Webb Johnson, G., & Bridgest, S. T. (2003). The effects of African American movement styles on teachers perceptions and reactions. The Journal of Special Education, 37(1), 49 57.
120 Neill, J. R., & Kniskern, D. P. (Eds.). (1982). From psyche to system: The evolving therapy of Carl Whitaker. New York: Guilford. Organisation for Economic Cooperation and Development. (2003). The PISA assessment framework: Mathematics, reading, science, and problem solving knowledge and skills. Paris: Organisation for Economic Cooperation and Development. Pagani, L., Boulerice, B., Vitaro, F., & Tremblay, R. E. (1999). Effects of poverty on academic failure and delinquency in boys: A change and process model approach. Journal of Child Psychology and Psychiatry, 40(8), 12091219. Palardy, J. M. (1998). The effects of teachers expectations on childrens literacy development. Reading Improvement, 35(4), 184 186. Perry, R. P., Stupnisky, R. H., Daniels L. M. & Haynes, T. L. (2008). Attributional (explanatory) thinking about failur e in new achievement settings. European Journal of Psychology of Education, 23(4), 459 475. Peter, K., Horn, L., & Carroll, C. D. (2005). Gender differences in participation and completion of undergraduate education and how they have changed over time. Washington, D C: National Center for Education Statistics. Planty, M., Bozick, R., & Ingels, S. J. (2006). Academic Pathways, Preparation, and PerformanceA Descriptive Overvie w of the Transcripts from the High School Graduating Class of 2003 04 (NCES 2007 316). U.S. Department of Education, National Center for Education Statistics. Washington, D C: U.S. Government Printing Office. Planty, M., Hussar, W., Snyder, T., Provasnik S., Kena, G., Dinkes, R., et al. (2008). The condition of education 2008 (NCES 2008 031) National Center for Educational Statistics, Institute of Education Sciences, U.S. Department of Education. Washington, DC. Preckel, F., H olling, H., & Vock, M. ( 2006). Academic underachievement: Relationship with cognitive motivation, achievement mot ivation, and conscientiousness. Psychology in the Schools, 43(3), 401 411. Prel ow, H. M. & Loukas, A. (2003). The role of resource, protective, and risk factors on academic achievement related outcomes of economical ly disadvantaged Latino youth. Journal of Community Psychology, 31 (5), 513529. Rodriguez, M. C. & Morrobel, D. (2004). A review of Latino youth development research and a call for an asset orientation. H ispanic Journal of Behavioral Sciences, 26(2), 107 127.
121 Saenz, V. B. & Ponjuan, L. (2009). The vanishing Latino male in higher education. Journal of Hispanic Higher Education, 8(1), 54 89. Sax, L. (2007). Boys Adrift: The five factors driving the growing epidemic of unmotivated boys and underachieving young men. New York: Basic Books. Schwartz, W. (2002). Helping Underachieving Boys Read Well and Often. ERIC Clearinghouse on Urban Education New York, NY. ERIC Digest. ED467687 Sciarra, D., & Whitson, M. (2007). Predictive factors in postsecondary educational attainment among Latinos. Professional School Counseling, 10(3), 307 316. Sciutto, M. J., Nolfi, C. J., & Bluhm, C. (2004). Effects of child gender and symptom type on referrals for ADHD by elementary school teachers. Journal of Emotional and Behavioral Disorders, 12(4), 247 253. Shields, S. A. (2008). Gender: An intersectionality perspective. Sex Roles, 59(5 6), 301311. Skarbrevik, K. J. (2002). Gender differences among students found eligible for special education. European Journal of Special Needs Education, 17(2), 97 107. Skiba, R. J., Michael, R. S., Nardo, A. C., & Peterson, R. L. (2002). The color of discipline: Sources of racial and gender disproportionality in school punishment. The Urban Review, 34 (4), 317342. Smith, D. D. (2007). Introduction to Special Education: Making a Difference (6th ed.). Boston: Allyn & Bacon. Smith, E. (2003). Understanding underachievement: An investigation into the differential attainment of secondary school pupils. British Journal of Sociology of Education, 24, 575586. Smith Adcock, S., Daniels, M., Lee, S., Villalba, J., & Indelicato, N. (2006). Culturally responsive school counseling for Hispanic/Latino students and fa milies: The need for bilingual school counselors. Professional School Counseling, 10 (1), 92 101. Snipes, J., William s, A., & Petteruti, A. (2006). Critical trends in urban education: Sixth survey of Americas Great City Schools. Washington, D C: Council of the Great City Schools. Somers, C. L., Owen s, D., & Piliawsky, M. (2008). Individual and social factors related to urban African American adolescents school performance. The High School Journal, 91(3), 111.
122 Spence, J.T. & Helmreich, R.L. (1983). Achievement related motives and behaviors. In J. T. Spence (Ed.), Achievement and Achievement Motives: Psychological and Sociological Approaches (pp. 774). San Fransisco: Freeman. Stanard, R. (2003). High school graduation rates in the United States: Implications for the counseling profession. Journal of Counseling & Development, 81(2), 217221. Sto ne, C.B. & Dahir, C.A. (2006). The Transformed School Counselor. Boston: Houghton Mifflin C ompany. Taylor, A.Z. & Graham, S. (2007). An examination of the relationship between achievement values and perceptions of barriers among low SES African American and Latino students. Journal of Educational Psychology, 99(1), 52 64. The College Board. (2 008). The SAT w riting section. Retrieved September 15, 2009 at http://www.collegeboard.com/prod_downloads/sat/thesat writing section.pdf The College Board. (2010). The Educational Crisis Facing Young Men of Color. Retrieved September 2, 2010 at http://advocacy.collegeboard.org/preparationaccess/educational cri sis facing young mencolor Thompson, B. (2004). Exploratory and Confirmatory Factor Analysis: Understanding Concepts and Applications. Washington, D C: American Psychological Association. Tinto, V. (1987). Leaving College: Rethinking the Causes and Cures of Student Attrition. Chicago: University of Chicago Press. Trusty, J. (2002). Effects of high school course taking and other variables on choice of science and mathematics majors. Journal of Counseling and Development, 80(4), 464 474. Trusty, J., & Brown, D. (2005). Advocacy competencies for professional school counselors. Professional School Counseling, 8(3), 259 265. Turkheimer, E., Haley, A., Waldron, M., DOnofrio, B., & Gottesman, I. (2003). Socioeconomic status modifies hereditability of IQ in young children. Psychological Science, 14(6), 623628. U. S. Census Bureau. (2008). An older and more diverse nation by midcentury. Retrieved September 27, 2009 at www.census.gov/Press Release/www/releases/archives/population/012496.html U. S. Department of Agriculture. (2009). National school lunch program. Retrieved October 1, 2009 at http://www.fns.usda.gov/cnd/Lunch/
123 U.S. Department of Education. (2001). No Child Left Behind Retrieved October 1, 2009, from http://www.ed.gov/nclb/landing.jhtml. Van De Gaer, E., Pustjens, H., Van Damme, J., & De Munter, A. (2007). Impact of attitudes of peers on language achievement: Gender differences. Journal of Educational Research, 101(2), 78 92. Van Houtte, M. (2004). Why boys achieve less at sc hool than girls: The difference between boys and girls academic culture. Educational Studies 30 (2), 159173. Varlas, L. (2005). Bridging the widest gap: Raising the achievement of black boys. Education Update, 47(8), 1 4. Walsh, M. E., Barret t, J. G., & DePaul, J. (2007). Day to day activities of school counselors: Alignment with new directions in the fiel d and the ASCA National Model. Professional School Counseling, 10(4), 370378. Wehmeyer, M. L. & Schwartz, M. (2001). Disproportionate representation of males in special education services: Biology, behavior, or bias? Education and Treatment of Children, 24(1), 28 45. Weiner, B. (1972). Theories of Motivation: From Mechanism to Cognition. Chicago: Rand McNally. West, A. & Pennell, H. (2003). Underachievement in Schools. London; New York: RoutledgeFalmer. West Olatunji, C., Shure L., Pringle, R., Adams, T., Lewis, D., & Cholewa, B. (2010). Exploring how school counselors position low income African American girls as mathematics and science learners. Professional School Counseling 13(3), 184 195. Wigfield, A. & Eccles, J. S. (2000). Expectancy Value theory of achievement motivation. Contemporary Educational Psychology, 25(1), 6881. Wigfield, A. & Eccles, J. S. (Eds.). (2002). Development of achievement motivation. San Diego: Academic Press. Wittmer, J. & Clark, M. A. (Eds. ). (2007) Managing your school counseling program: K 12 Developmental strategies (3rd ed. ). Minn eapolis, MN: Educational Media. Wong, C. A., Eccles, J. S., & Sameroff, A. (2003). The influence of ethnic discrimination a nd ethnic identification on African American adolescents school and socioemotional adjustment. Journal of Personality, 71(6), 11971232.
124 Woolley, M. E., Grogan Kaylor, A., Gilster, M. E., Karb, R. A., Gant, L. M., Reischl, T. M., & Alaimo, K. (2008). Neighborhood social capital, poor phy sical conditions, and school achievement. Children & Schools, 30(3), 133145.
125 BIOGRAPHICAL SKETCH Erin Leigh Oakley was born in Richmond, Virginia t o parents Rick and Pam Oakley. She was joined by a brother, Craig David Oakley, two years later, and was raised in this loving family of four in Midlothian, Virginia. She had an active childhood, always involved in school, church, and athletics. She attended J. B. Watkins Elementary School, Midlothian Middle School, an d Midlothian High School. After gra duating from high school she attended the University of Virginia in Charlottesvill e, where she double majored in psychology and French language and l iterature. Upon gr aduation from UVA she made plans to attend the University of F lorida for graduate studies in counselor education. She was awarded an A lumni Fellowship to pursue her M aster of E ducation and E ducation S pecialist degrees as well as her doctorate in the field of Mental Health Counseling. Erin graduated w ith h er M.Ed. and Ed.S. in 2005. Then, after completing her doctoral coursework and clinical work, Erin moved to Virginia for one year, during which time she married Robert Mark Knape, M.D. and worked at Christopher Newport Universitys counseling center. Now Erin Oakley Knape, she re turned to Gainesville to resume work on her doctoral degree, and expects to graduate from the Universit y of Florida with her Ph.D. in counselor education in December of 2010. Erin and her husband, Rob currently reside in Gai nesville, Florida with their cat, Bear. The couple is expecting their first child, a son, in the fall of 2010. During rare free time, Erin enjoys reading, swimming, going to the beach, traveling, and spending time with friends and family.