Translating Knowledge From an Evaluation Report
Necessary, But Not Sufficient: The McKinney- Vento Act and Academic Achievement
in North Carolina George Hendricks and William Barkley
The McKinney-Vento Homeless Education Assistance Improvements Act of 2001 (MCKV), which provides funds to local educational agencies (LEAs), is almost a decade old, yet no evaluations of its academic effectiveness have been reported. Using a systems theory framework, the authors answer the question of whether homeless students in grade 6 from LEAs that received MCKV funding scored better in reading comprehension and mathematics on end-of-grade (EOG) test scores than did those from LEAs that did not receive MCKV funding. Data from 2006 and 2007 were provided by the North Carolina Department of Public Instruction. About 20 percent of the state's LEAs received MCKV grants, which created a treatment group (funded LEAs) and a control group (nonfunded LEAs). Using untreated control group designs with matched pretests (grade 5 EOG test scores) and posttests (grade 6 EOG test scores), 2 x 2 analyses of variance with repeated measures failed to reject the nuU hypotheses. The authon did not support the hypotheses that MCKV grants improved the academic achievement of homeless students. There was no significant difference in the funded and nonfunded EOG test scores.
KEY WORDS; academic achievement; end-of-grade test scores; homeless education; homelessness McKinney- Vento Act
C ongress has been interested in improving the education of homeless students since 1987, particularly since the McKinney-
Vento Homeless Education Assistance Act of 2001 (P.L. 107-110) (MCKV), the major federal investment in homeless education. A hterature search concurred with Müler's (2009) assertion that there is a shortage of empirical analysis of the MCKV. The MCKV has never been the subject of a statewide empirical evaluation. Müler's obser- vation of tbe lack of analysis of MCKV highlight- ed the need for the present study—arguably tbe fint empirical evaluation of tbe academic effec- tiveness of MCKV. Our data suggest tbat the MCKV did not significantly improve the academ- ic performance of students in North Carohna. We also saw that although the MCKV provides solu- tions to some problems faced by bomeless stu- dents—including transportation, ease of registration, scbool suppUes, and some health issues—in this study it did not result in significant academic improvements for homeless students.
JuhaneUe (2008) noted that thousands of school chüdren in the United States experience
homelessness each day. The negative impact of homelessness on the hves of chüdren has been weU documented. Compared witb conventionaUy boused chüdren, homeless chüdren have poor bealtb and limited preventive bealtb care support. Hopper (2003) and Wong, Salomon, Tbisde- EUiott, TaUarita, and Reed (2004) reported that homeless chüdren experience more types of ulness for longer periods of time and with more harmful consequences than conventionaUy housed stu- dents. Simüarly, the negative impact of homeless- ness on psychological and mental health issues has been reported by Swick (2006) and Taylor, Stut- taford. Broad, and Vostanis (2006).
School social worken are particularly interested in the impact of homelessness on students' aca- demic achievement. Israel, Urberg, and Toro (2001); Jozefowicz-Simbeni (2003); and Rafferty (1998) reported that compared with conventional- ly housed students, homeless students have lower achievement test scores, poorer grades, more grade retentions, and a higher incidence of school dropout. Hendricks and Barkley (2011) compared scores on the end-of-grade (EOG) tests of
doi: 10.1093/cs/cds007 « 2012 National Association of Sociai Wori<ers 179
homeless and conventionally housed grade 6 stu- dents in the state of North Carolina, Their exami- nation of four EOG tests (reading comprehension in 2006 and 2007 and mathematics in 2006 and 2007) supported the finding that conventionally housed grade 6 North Carolina students scored significantly better than did homeless grade 6 students.
The nation's homeless school-age population has increased by 41 percent over the past two school yean (National Association for the Educa- tion of Homeless Chüdren and Youth [NAEHCY], 2010), North Carolina experienced a 48 percent increase in homeless school-age chü- dren from the 2006-07 school year to the 2008- 09 school year (NAEHCY, 2010), This recent growth in the number of homeless students has intensified the need to provide effective programs for homeless students and has increased the need to evaluate the academic efficacy of such programs,
THE MCKINNEY-VENTO PROGRAM The fint major congressional step regarding edu- cation of homeless students was taken in 1987 with passage of the McKinney-Vento Homeless Assistance Act (P,L, 100-77) (more commorüy known as the Stewart B, McKinney Act), By en- acting this legislation, the federal govemment rec- ognized that it has a responsibility to assist in improving the educational experiences of home- less chüdren. This act provided programs for trans- portation, emergency food, housing, and emergency mental health care for homeless chü- dren. Services provided by the act were increased through amendments in 1988, 1990, 1992, and 1994 (Project Hope, 2008), A major revision oc- curred in 2001, when the Stewart B, McKinney Act was reauthorized as MCKV, This new act mandated the designation of a state coordinator for homeless students; made it easier for homeless students to register in new schools; and provided students with free lunch, clothing, school supplies, and other amenities.
One of the key concems of education is aca- demic success. But there are certainly other con- cems, such as political, vocational, economic, and moral ones. It might be argued that MCKV is valuable in areas beyond academics. The purpose of this study, however, was to determine the impact of MCKV on academic success. As early as
1990, Congress recognized that the purpose of the Stewart B, McKinney Act was not simply to enroll homeless chüdren in school, "but to promote their academic success in public schools" (Project Hope, 2008, p, 1),
LACK OF EVALUATION OF THE MCKV A search of the literature on homelessness (Aca- demic Search Premier, PsycARTICLES, Social Service Abstracts, and Sociological Abstracts) re- vealed a large amount of scholarly research on homelessness in general and on the impact of homelessness on some aspects of education, Andenon (2003) observed that the volume of homeless literature was almost unmanageable. However, no empirical studies were found com- paring the academic achievement of homeless stu- dents in local educational agencies (LEAs) receiving MCKV support with homeless students in unsupported LEAs,
Beginning with the 2002-03 school year, the federal government required states to provide aca- demic achievement data on homeless students. Al- though reading comprehension and mathematics scores for homeless chüdren across the nation have improved (National Center for Homeless Educa- tion, 2008), it has not been possible to determine from these data the role of MCKV in the improvement,
Müler (2009) pointed out that the MCKV "is almost univenally lauded by homeless educators and advocates as a highly beneficial asset for stu- dents and families" (p, 8), Müler noted that the MCKV has hberaHzed the definition of homeless- ness, expanded each LEA's transportation duties, increased the number of homeless students being tested, and contributed to improving the reading comprehension and mathematics scores of home- less students. We agree with Miller regarding the positive contributions of MCKV, but we could find no empirical studies that addressed our ques- tion concerning the impact of the MCKV on aca- demic achievement.
The EOG test scores that states are required to report involve all homeless students, but not all homeless students benefit from the MCKV, as was the case in North Carolina in the 2006-07 school year. The present study is unique in that we did not consider the homeless students a ho- listic group; instead, we compared the educa- tional achievement of MCKV-supported homeless
180 Children & Schools VOLUME 34, N U M B E R 3 JULY 2012
students with the achievement of homeless stu- dents not receiving the benefits of the MCKV. Our literature search revealed that no such com- parison had been done previously. Such a com- parison was necessary to determine with confidence the role of the MCKV in academic achievement.
As part of the literature review, we sought the advice of people who were especially knowledge- able about homeless education in general and the MCKV program in particular. First, one of us (George Hendricks) spoke with the person who supervised the MCKV program at the U.S. Department of Education. When Hendricks asked about peer-reviewed evaluations of the aca- demic effectiveness of the MCKV, the supervisor replied, "If you find anything, let me know" (G. Rutkin, personal communication. May 25, 2007).
Hendricks then spoke with the person who for five years had the primary responsibility in the North Carolina Department of Public Instruction (NCDPI) (2002) for implementing the MCKV. She was arguably the most informed person in North Carolina regarding the effectiveness of the MCKV in North Carolina. When she was asked where we might look for scholarly, peer-reviewed assessments of the MCKV, she, like the Depart- ment of Education official, stated that she was not aware of any peer-reviewed empirical studies on the academic effectiveness of the MCKV program (D. McHenry, personal communication, October 28, 2007).
Is this lack of empirical evaluation regarding the academic impact of the MCKV important? Is the academic achievement of homeless students in funded LEAs compared with that of homeless stu- dents in unfunded LEAs an important question? It might be argued that the services that the MCKV provides to homeless students are sufficient in themselves, regardless of the academic issues. For example. Markward and Biros (2001) estimated that the Stewart B. McKinney Act increased the school attendance of homeless students by 17 percent. The argument could be made that this increase is sufficient justification for the MCKV, regardless of whether it has improved academic achievement.
We recognize and appreciate the contributions of the MCKV to the educational process. We un- derstand, however, that a fundamental purpose of education is academic. Improved support services
that do not lead to improved academic achieve- ment would signal that the MCKV is a valuable but insufficient program. It is important for Con- gress and educational leaders to understand the ac- ademic efficacy of this major program.
DATA COLLECTION A N D ANALYSIS
The absence of an empirical evaluation of MCKV's impact on academic success constitutes a significant gap in the literature. Our response was to investigate one component of this absence— namely, MCKV's effect on homeless North Caro- lina grade 6 students. Our decision to study the 2005-06 and 2006-07 school years (test dates: spring 2006 and spring 2007) was influenced by the person who has directed homeless education in North Carolina for many yean. From Deborah McHenry's (personal communication, October 28, 2007) experience, 2006 was the first year in which North Carolina had useful quantitative data on homeless education. The most recent year that data were available at the time of our study was 2007.
We used archival data from two groups that had been estabhshed prior to the study. In North Carolina in 2006 and 2007, only about 20 percent of the LEAs, in most cases counties, received MCKV funding. The two groups resulted from a kind of self-selection. Each LEA chose to apply, or not to apply, for grants made available from the state under the MCKV; in short, approxi- mately 20 percent of the LEAs in North Carolina chose to receive MCKV grants, and approximate- ly 80 percent did not. This difference created the possibility of a funded experimental group receiv- ing MCKV support and an unfunded control group not receiving MCKV funding. The anomaly of having more unflinded LEAs but a smaller unfiinded sample size is revealed in this ar- ticle's tables. This anomaly results from the fact that the more populated LEAs were overrepre- sented in the funded 20 percent. In fact, all of the large and wealthier urban LEAs were represented in the funded group. Many of the smaller, rural LEAs had not gathered and reported data on homeless students, and most of the smaller LEAs did not apply for the grants. This imbalance re- sulted in an urban bias in this study.
The NCDPI (2002) provided the grade 6 EOG test scores for mathematics and reading compre- hension for the test in the spring of 2006 and the
HENDRICKS AND BARKLEY I Necessary, But Not Sufficient: McKinney-Vento and Academic Achievement 181
spring of 2007. The grade 5 scores (that is, 2005 for the 2006 test and 2006 for the 2007 test) were provided and served as a baseUne. Our study was guided by one research question: Do homeless students from LEAs that receive MCKV funding achieve higher EOG reading comprehension and mathematics scores than do students from LEAs that are not funded? The analysis compared the change from the pretest (grade 5) to the posttest (grade 6) scores of homeless students in the LEAs receiving funding with the EOG scores of home- less students in the LEAs that were not funded. Separate tests were done for reading comprehen- sion and mathematics for 2006 and 2007.
Analysis was Umited to students who had both grade 5 pretest scores and grade 6 posttest scores. This practice of requiring a matching pretest base- line had both positive and negative effects. Posi- tively, it meant that all indî vidual posttest student scores were matched to the pretest scores of the same students. This provided a useful baseUne. Negatively, the process significantly reduced the number of scores avaUable, because all unmatched cases were omitted. The absence of a pretest score was more common than the absence of a posttest score, but instances of both occurred. Approxi- mately 26 percent of the 2006 data and 23 percent of the 2007 data were unmatched and were subse- quendy not used in the study. This loss of data was regrettable, but it could not be avoided without loss of the baseline.
As the result of confidentiality and public rela- tions concems, our data use agreement with the NCDPI (2002) sharply limited the demographic detaU of the sample. For example, it was not pos- sible to compare the students who were dropped from the sample with those who continued. We conjectured, however, that at least three faetón contributed to the missing data: (1) A student could have been homeless in one year (pretest or posttest) but not in the other. (2) A student could have been homeless in both the pretest and post- test yean but could have moved to a different LEA between the pretest and posttest yean, a result not uncommon among highly mobile homeless students. (3) Attendance among home- less students is historically low, and poor atten- dance could have resulted in a student's absence from school on the date of either the pretest or the posttest.
One independent variable was the test year, 2006 or 2007, and the other was the funding category, funded or nonfunded. The dependent variables were ratio-level EOG scores for read- ing comprehension and mathematics. A 2 x 2 repeated-measures analysis of variance (ANOVA) was used to determine whether there was a signif- icant difference between the EOG scores for the funded and nonfunded groups. In aU four cases, histogram plots confirmed the normaUty of the data, and Levene's tests confirmed the homogene- ity of variances.
RESULTS
Descriptive statistics for reading comprehension EOG scores, comparing the same students from grade 5 in 2005 and grade 6 in 2006, are displayed in Table 1. A 2 x 2 repeated-measures ANOVA found no significant {p < .05) interaction between grade and funding [F(l, 337) = 0.026, jj=.872]; therefore, we tested the main effect of grade and the main effect of funding. The difference between the total pretest (grade 5) and total post- test (grade 6) reading scores combined across funding states (that is, the main effect of grade) was not significant [F(l, 337) = 1.14, j? = .286]. In addition, the difference between students from funded LEAs and students from nonfunded LEAs combined across pretests and posttests (that is, the main effect of funding) was not significant [F(l, 337) = 2.81, p = .O94]. There was no significant difference between funded and nonfunded scores.
Table 1: Reading Comprehension: 2006 Posttest versus 2005 Pretest
Funding status
Pretest (Grade 5, 2005)
Posttest (Grade 6, 2006) Total
Funded M
SD
Number Nonfiinded
M
SD Number
Total M
SD
Number
252.55 10.33
250
254.37 11.39 89
253.46 11.11
339
253.44
12.85 250
255.02 12.63 89
254.23 10.24
339
252.99 9.09
500
254.69 12.26
178
253.84
10.67 678
Note: The range of scores for 2006 was 137 through 281, with higher scores indicating greater reading proficiency.
182 Children & Schools VOLUME 34, N U M B E R 3 JULY 2012
Table 2: Reading Comprehension: 2007 Posttest versus 2006 Pretest
Funding status
Pretest (Grade 5, 2006)
Posttest (Grade 6, 2007) Total
Funded M
SD
Number Nonftinded
M
SD
Number Total
M
SD
Number
345.13 19.89
249
345.56 19.50 86
345.34
19.69
335
348.42
8.19
249
348.36 7.98
86
348.39 8.08
335
346.87 14.04
498
346.49 13.74
172
346.86 13.89
670 Note: The range of scores for 2007 was 234 through 337. with higher scores indkat. ing greater reading proficiency.
Descriptive statistics for reading comprehension EOG scores, comparing the same students from grade 5 in 2006 and grade 6 in 2007, are displayed in Table 2. A 2 x 2 repeated-measures ANOVA found no significant {p < .05) interaction between grade and funding [F(l, 333) = .031, p=.86O]; therefore, we tested the main effect of grade and the main effect of funding. The difference between the total pretest (grade 5) and total post- test (grade 6) reading scores combined across funding states was significant [F(l, 333) = 4.94, p=.O27]. The difference between students from funded LEAs and students from nonfunded LEAs combined across pretests and posttests was not sig- nificant [F(l, 333) = 0.020, p=.88S]. There was no significant difference between funded and nonfunded scores.
Descriptive statistics for mathematics EOG scores, comparing the same students from grade 5 in 2005 and grade 6 in 2006, are displayed in Table 3. A 2 x 2 repeated-measures ANOVA found no significant {p < .05) interaction between grade and funding [F(l, 187) = 0.273, p=.6O2]; therefore, we tested the main effect of grade and the main effect of funding. The difference between the total pretest (grade 5) and total post- test (grade 6) reading scores combined across funding states was significant [F(l, 333) = 12.43, p=.OOl]. The difference between students from funded LEAs and students from nonfunded LEAs combined across pretests and posttests was not sig- nificant [F(l, 187) = 0.008, p=.93O]. There was
Table 3: Mathematics: 2006 Posttest versus 2005 Pretest
Funding Pretest Posttest status (Grade S, 2005) (Grade 6, 2006) Total Funded
M SD Number
Nonfunded
M
SD
Number Total
M
SD
Number
252.69 6.91
154
252.31 8.70
35
252.62
7.25
189
254.24 7.14
154
254.40 7.82
35
254.27 7.25
189
253.46 7.02
308
253.35 8.26
70
253.41 7.64
378 Note: The range of scores for 2006 was 100 through 276. with higher scores indicat- ing greater mathematics proficiency.
no significant difference between funded and nonfunded scores.
Descriptive statistics for mathematics EOG scores, comparing the same students from grade 5 in 2006 and grade 6 in 2007, are displayed in Table 4. A 2 x 2 repeated-measures ANOVA found no significant {p < .05) interaction between grade and funding [F(l, 335) = 0.150, p = .618]; therefore, we tested the main effect of grade and the main effect of funding. The difference between the total pretest (grade 5) and total post- test (grade 6) reading scores combined across funding states was significant [F(l, 335) = 8.25, p=.004]. The difference between students from funded LEAs and students from nonfunded LEAs combined across pretests and posttests was not sig- nificant [F(l, 335) = 0.113, p=.737]. There was no significant difference between funded and nonfunded scores.
Homeless students from the North Carolina LEAs that received MCKV funding did not dem- onstrate significantly different achievement in reading comprehension or mathematics than did homeless students from LEAs that did not receive MCKV funding. Homeless students did demon- strate small gains over time in three of four tests, but that was for all homeless students combined across the funded and the nonfunded LEAs. Cohen's (1969) method was used to evaluate effect size (ii<.23). Cohen classified this effect size as small, which was another way of conclud- ing that there was no significant difference
HENDRICKS AND BARKLEY INecessary, But Not Sufficient: McKinney—Vento and Academic Achievement 183
Table 4: Mathematics:
Fundiñg^PI
Funded M
SD
Number Nonfunded
M
SD
Number Total
M
SD Number
2007 Posttest versus 2006 Pretest
345.18
19.79 252
345.22
14.56
85
345.20 19.68
337
348.15 9.70
252
349.13 7.66
85
348.64
8.68
337
346.66
14.75 504
347.18 13.62
170
346.92
14.18 674
Note: rhe range of scores for 2007 was 235 through 377, with higher scores indicat-
ing greater mathematics proficiency.
between tbe experimental and control groups. Tbe absence of interaction effects for all four tests indicated tbat tbe pre- to posttest score changes were similar for homeless and conventionally housed students. For all eight groups, there was an increase of this score from pre- to posttest.
CONCLUSION
The primary focus of this study was determining whether MCKV funding improved the academic performance of homeless students. We concluded that although the MCKV as it is currently struc- tured makes valuable contributions to certain aspects of education, it fails to make a significant positive impact on the academic achievement of homeless students. The hypothesized association between tbe MCKV and improved academic achievement of homeless students was not demonstrated.
IMPLICATIONS
Our study failed to show that MCKV funding significandy improved the EOG scores of home- less students. The program, however—even at its current, moderately funded level—achieved the important supportive goal of increasing school at- tendance Qeynes, 2002; Kerbow, Azcoitia, & Buell, 2003; Markward & Biros, 2001).
Our study only empirically validated tbat the MCKV did not work effectively in improving the academic achievement of students in North Caro- lina, a state where few otber educational support
services for homeless students are available. A valuable and additional contribution to under- standing the contribution of tbe MCKV to aca- demic achievement would be an empirical study similar to ours conducted in a state witb a more extensive homeless-supportive environment.
Each LEA in North Carohna determines how the 18 approved uses of the MCKV subgrant funds are realized (National Center for Homeless Education, 2008). Some of these are probably more effective than others, but given the data available, it is impossible to identify wbicb ap- proaches might be more valuable in improving the academic achievement of homeless students. A further useful study would be an evaluation of the impact of the different uses of MCKV funds.
More resources for the MCKV would probably help to make MCKV more effective, although more resources in an unmodified program are not a guaranteed panacea. MCKV as currently struc- tured provides htde training or preparation for social workers, teachers, and counselors to belp them address the special needs of homeless stu- dents. Although there are no easy solutions, it is interesting that the special classroom needs of homeless students have received so httle MCKV support. Support for transportation, ease of regis- tration, and physical and mental health are valu- able, but the available services, hke MCKV in general, have an extra-classroom focus.
We agree with advocates for the homeless (for example, NAEHCY, 2008; Wong-et al., 2004) that the funds currently allocated to support homeless students are insufficient. Advocates for the homeless population have made a strong case for increased funding. We beheve tbat tbeir efforts are wortby of attention and support. Tbis empiri- cal study strengthens the case by NAEHCY and Wong et al. tbat more needs to be done. In short, MCKV is an underfunded program that is impor- tant but inadequate. We hypothesize that in envi- ronments where strong programs support homeless famüies, tbe MCKV program may provide the additional impetus needed for signifi- cant academic improvement. In addition, future studies need to place more emphasis to controUing for other factors such as rural and urban differenc- es, gender, race, and ethnicity. A thorough study needs to be conducted to address the large numbers of students who fail to meet botb pretest and posttest results. It was not possible to acbieve
184 Children & Schools VOLUME 34, N U M B E R 3 JULY 2012
such goals in this preliminary study because of the limited amount of information that the NCDPI (2002) was wuhng and able to release.
At least two other issues hmited the quality of the data. The fint issue was the data's lack of com- pleteness; many small mral counties did not coüect and report the needed data on homeless students, so the nonfunded n avaüable for analysis was low. The second issue involved confidentiali- ty issues that prevented us from tracking mobüe students who moved from one LEA to another.
Although we have concluded that a basics-only MCKV program did not achieve the desired aca- demic goals, we are confident that an enhanced MCKV could contribute to a significant improve- ment in the academic achievement of homeless students. The lack of academic effectiveness of the current MCKV subgrant process highlights the need for further study and amendments to the structure and level of support for MCKV, S
REFERENCES Anderson, I, (2003), Synthesizing homeless research:
Trends, lessons and prospects. Journal of Community & Applied Social Psychology, 13, 197-205,
Cohen,}, (1969), Statistical power analysis for the behavioral sciences. New York: Academic Press,
Hendricks, G,, & Barkley, B, (2011), Educational conse- quences of homelessness: The effects of homelessness on end-of-grade tests of grade 6 North Carohna stu- dents. School Social Work Journal, 36(1), 71-94,
Hopper, K, (2003), Reckoning with the homeless. Ithaca, NY: Cornell University Press,
Israel, N,, Urberg, K,, & Toro, P, A, (2001, August), Aca- demic performance and maladjustment in housed and home- less adolescents. Paper presented at the 119th Annual Convention of the American Psychological Associa- tion, San Francisco,
Jeynes, W, (2002), Divorce, family structure and the academic success of children. Binghamton, NY: Haworth Press,
Jozefowicz-Sinibeni, D,M,H, (2003, Apd]), Why do they leave? Why do they stay? Qualitative and quantitative per- spectives on high school dropout. Poster session presented at the biennial meeting of the Society for Research in Child Development, Tampa, FL,
Julianelle, P, F, (2008), The McKinney-Vento Act and children and youth awaiting foster care placement: Strategies for im- proving educational outcomes through school stability. Min- neapolis: National Association for the Education of Homeless Chñdren and Youth,
Kerbow, D,, Azcoitia, C , & Buell, B, (2003), Student mo- bility and local school improvement in Chicago, Joumal of Negro Education, 72, 158-164,
Markward, M, J,, & Biros, E, (2001), McKinney revisited: Implications for school social work. Children & Schools, 23, 182-187,
McKinney-Vento Homeless Assistance Act, P,L, 100-77, 101 Stat, 482 (1987),
McKinney-Vento Homeless Education Assistance Im- provement Act OÍ2001, P,L, 107-110, 115 Stat, 1989 (2002),
Miller, P, M, (2009, November), An examination of the McKinney-Vento Act and its influences on the
homeless education situation. Educational Policy, 25, 424-450,
National Association for the Education of Homeless Chil- dren and Youth, (2010, July), A critical moment: Child & youth homelessness in our nation's schools. Retrieved from http://www,naehcy,org/dl/crit_mom,pdf
National Center for Homeless Education, (2008), Analysis of 2006-2007 federal data collection and three year compar- ison. Retrieved from http://www,ed,gov/Programs/ homeless/index, h tml
North Carolina Department of Public Instruction, (2002), Standard course of study and grade level competencies. R e - trieved from http://www,dpi,state,nc,us/curriculum/ ncscos
Project Hope, (2008), History of the McKinney Act: A report of the National Center for Homeless Children and Youth. Retrieved from http://www,wm,edu/hope/national/ mckinney/mckinney,htnil
Rafferty, Y, (1998), Meeting the educational needs of homeless children. Educational Leadership, 55(4), 38-52,
Swick, K, J, (2006), Helping homeless famihes overcome barriers to successful functioning. Early Childhood Ed- ucation Journal, 33, 195-200,
Taylor, H,, Stuttaford, M,, Broad, B,, & Vostanis, P, (2006), Why a roof is not enough: The characteris- tics of young homeless people referred to a designated mental health service. Joumal of Mental Health, 15, 491-501,
Wong,J,, Salomon, A,, Thistle-Elliott, L,, Tallarita, L,, & Reed, S, (2004), The McKinney-Vento Homeless Assistance Act—Education for homeless children and youths program: Turning good law into effective ed- ucation, Georgetown Joumal on Poverty Law & Policy, Ji, 283-319,
George Hendricks, PhD, C-SSWS, is associate professor, director of the director of the BSW program, and chair. School of Social Work, Methodist University, 5400 Ramsey Street, Eayetteville, NC, 28311, e-mail: ghendricks@methodist. edu. William Barkley, PhD, is contributing faculty member. School of Health Sciences, College of Health Scienc- es, Waiden University, Minneapolis, MN.
Original manuscript received October 8, 2010 Final revision received March 10, 2011 Accepted May 2, 2011 Advance Access Publication September 20, 2012
TRENDS & RESOURCES
Trends & Resources presents currentpractice trend information accom- panied by highlights of relevant books, curriculums, films, and other practice aids for school social workers and their colleagues. The column is co-edited by an academic researcher and a practitioner to help bridge the gap between the latest weü-researched tools, current policy and practice issues, and the field. The journal does not accept unsolicited reviews for this column.
HENDRICKS AND BARKLEY ¡Necessary, But Not Sufficient: McKinney-Vento and Academic Achievement 185
Copyright of Children & Schools is the property of National Association of Social Workers and its content may
not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.