Introduction and Theoretical Backgroung 

Computers and similar devices such as cell phones and video games are a staple of modern existence and their continued advancement into every aspect of our lives shows no signs of abating. Reading instruction falls under this umbrella of computer use in all aspects of life. Many classrooms now feature computers in the class and or have computer labs located in the school which the students can access. There is a wide array of computer games created to take advantage of the above facts. The makers of these games all claim that they benefit children in the process of acquiring the ability to read. Computer programs are developed for many age ranges and for different aspects of reading development such as building phonemic awareness and comprehension skills. The following research review will focus on computer programs created for elementary aged students (grades K-5) and with a focus on studies that deal with the efficacy of programs to help children who have been identified as having difficulties reading. Intervention in the lower grades is critical considering studies showing Mathew Effects in reading where the “rich” students who have stronger earlier skills get “richer” as they progress into later grades and adulthood vs. the “poor” students who have weaker skills. (Stanovich, K. (1986).

            Researchers have identified phonological awareness as the core element of reading acquisition. (Pokorni, JL, Worthington, CK, & Jamison, PJ. (2004) Thus, all of the studies discussed in this research review deal with games that contain a phonological awareness component. Many of the computer programs discussed also have components that teach strategies of other decoding skills such as segmentation and blending. Due to the large amount of computer assisted intervention (CAI) programs available, this research review will focus on some of the best known programs such as Fast Forword and ABRACADABRA in addition to studies done on computer assisted intervention programs that are similar to many others that are on the market. 

Review of Research

          ABRACADABRA ( A Balanced Reading Approach for all Canadians Designed To Achieve Best Results For All) is a free online program that can be used in the classroom as well as at home. (2008. ABRA description) Savage, Deault, and Abrami conducted a randomized controlled trial study of the ABRACADABRA reading intervention program in grade 1 win order to research whether or not the program improves a range or reading skills and related skills compared to a control group. The quantitative study was conducted using 144 students from 13 first grade classroom in both suburban and urban school districts with children that came from a diverse set of socioeconomic backgrounds and languages spoken at home. The children in the experimental group used the ABRACADABRA program in school over the course of a school year for 20 minutes 4 times a week. A wide range of the students’ readings skills where measured on pre-tests and post tests using Woodcock-Johnson Test of Achievement, the Elison subtest from the Comprehensive Test of Phonological Processing, and the Group Reading Assessment and Diagnostic Evaluation (GRADE). The program content was based on what reviews have shown to be effective reading intervention for phonic and letter skills, reading fluency, and reading comprehension. An adult trained to use the program was assigned to groups of 4 students as they used the program in order to guide them through the program when necessary. Savage, RS, Deault, L, & Abrami, Ph. (2009).

          The results of the study showed significantly greater increases in phonological awareness, listening comprehension, and reading comprehension, on initial post tests and on post test administered seven months after the study had concluded. While the researchers discussed their finding they pointed out that ABRACADABRA showed larger gains on a variety of tasks than other similar programs and speculated that part of the reason for this might have been the inclusion of adults to guide children while they were using the program in the study. (Savage, RS, Deault, L, & Abrami, Ph. (2009).)

          Similar to the ABRACADABRA CAI program, are the Fast ForWord and Earobics programs. The efficacy of those two CAI programs was researched by Pokorni, Worthington, and Jamison. However in this study a third program Lindamood and Phonics Sequencing Program (LiPS) which does not have a computer assisted intervention component and is delivered by a trained instructor to a small group of children was also measured. Consequently, it was possible to measure the efficacy of the two computer based programs not only against a control group that did not receive any intervention but also against a group that received an intervention that was not computer based. (in this case LiPs)

            The study was conducted on 62 students who were separated into the three different intervention groups who were either seven or eight years of age, on an IEP, were one year behind in their reading level, and were from English speaking homes. The vast majority of the students in the study where African American (77%) with the other percentages being Caucasian (11%) Hispanic (7%) and Asian (3%). Phonological Awareness Tests and The Woodlock Language Proficiency Tests where used as pre and post tests measures with the post test being conducted four weeks after the intervention ended. The results of the study showed the largest increases in abilities of phonemic segmentation tasks where for the Fast ForWord and Earobics groups which both had similar increases. The two computer based intervention did not have significant increases in blending tasks which is the area where students receiving the LiPS intervention showed the greatest gains. Participants in all three interventions showed increases in post test performances compared to students that did not receive any intervention. However, the students who all initially where a year behind fellow students in reading level, still showed deficits compared to other students in more normative reading ranging that where not involved in the study. ( Pokorni, JL, Worthington, CK, & Jamison, PJ. (2004)      

          Of the programs researched in the above study, Fast ForWord is the most heavily marketed to schools. Fast ForWord was featured in a segment on NBC Nightly News and over 65,000 students in 2400 school districts have received training using one or more of the software products. (Troia , GA, & Whitney , SD. , 2003).  In addition the cost of Fast ForWord has been reported to be $38,000 per 50 students for a school district. (2009) Metiri Group Fastforward.) Hence this research review will go into deeper look to determine how effective Fast ForWord is. Hook, Macaruso, and Jones conducted a longitudinal study to determine the efficacy of Fast ForWprd as it compared to more standard intervention such as Orton Gilligham training sessions done by an actual tutor one on one with a child. In addition, they also compared to Fast ForWord and Orton Gillingham to a longitudinal control group made up of similar students.

            The Hook, Macaruso, and Jones study found their participants for the study in a different manner than any other study mentioned in this review; they placed an add in a Boston daily newspaper. Out of the parents who responded to the add 11 students where chosen to receive the Fast ForWord training program. Then similar groups of 11 children were found in order to receive the Orton Gillngham training sections and to be part of the control group. It should be noted that all the participants in the study ended up being Caucasian and from middle to upper middle class families; this again sets this study as unique from other conversed about in this review because nearly all of the others focus on children from lower socioeconomic backgrounds and or participants who are from a minority group.

            In the Hook, Macaruso, and Jones study the three groups were to measured 4 times: before the study began, at the completion of the study at the end of the summer, a year after the study was complete and two years after the study was complete. The range of tests used to measure students performance in many aspects of reading development included the Lindomood Auditory Conceptualization Test, The Woodcock Reading Mastery Test, Woodcock-Johnson Test of Language Abilities, and the Test of Language Development (TOLD). The actual study was conducted over 2 months during summer vacation with the Fast ForWord group receiving 5 twenty minute computer activities daily for a total of 100 minutes and the OG group receiving one on one remediation with a tutor for one hour a day and five days a week. (Hook, PE, Marcaruso, P, & Jones, S. , 2001).

            The results of the study showed that both FFW students and OG students made similar gains in phonemic awareness that were higher than the control group in the tests immediately after the programs where completed, but only the OG group made significant gains in word attack scores. Furthermore, the tests given 2 years later showed again similar results when FFW was compared to OG and to a control group in the areas of word attack, word identification, and passage comprehension. (Hook, PE, Marcaruso, P, & Jones, S. , 2001).

            The previous study mentioned seemed to show benefits when students used the Fast ForWord program at least when compared to a control group that did not receive and specific treatment program. Troia and Whitney conducted  a similar study in an attempt to ascertain whether or not Fast ForWord not only showed benefits for students, but whether those benefits where large enough to justify the claims made by Fast ForWord’s creators which where based on research conducted by the creators. Some of the major differences in Troia and Whitney’s study as compared to other studies about the effectiveness of Fast ForWord where that Troia and Whitney conducted their study in a rural school and they also surveyed parents of the participants to gauge their feeling as to whether or not FFW was a helpful tool for their children.

            The Troia and Whitney study included students in grades 1 through 6 from a rural Elementary School in the Puget Sound area. It should be noted, that this area is near a military base and thus the majority of students in the study had at least one parent in the military. Students where recommended to the program by their teachers because of low performance and the belief that they would benefit from intensive language based instruction. The demographics of the students in the experimental group where Caucasian 67%, Asian 11%, Hispanic, 8% African American and 3% Native American. The experimental and control groups were given pre and post test in order to measure the children’s abilities in 4 domains: phonological processing abilities, oral language competency, basic reading skills, and classroom behavior. The post test were given 10 weeks after the treatment program that consisted of 20 minute sessions with Fast ForWord each school day for 4 weeks for the experimental group; these sessions were conducted while the control groups where receiving typical language arts instruction in the classroom. This is another sphere of this particular study that was distinctive compared to others in this research review. Most of the studies attempting to find the benefits of computer assisted interventions took place in conjunction with regular language arts training in the classroom, but in this case the FFW intervention was done as a replacement for traditional in class instruction. (Troia , GA, & Whitney , SD. (2003 ).

            The FFW group achieved significantly greater gains than the control group in the areas of oral language, phonological awareness, and classroom behavior. The improvements were limited in scope however, For example, increases in blending tasks where significant on post tests for the experimental group vs. the control, but not in segmentation. On the other hand, the greatest increases between the experimental and the control group where seen for children who were the lowest performers in the experimental group vs. the lowest performers in the control group.  (Troia , GA, & Whitney , SD. 2003 )

            Now this review will focus more on the research that deal less with a specific intervention program such as Fast ForWord, and more on the general concept of computer assisted instruction (CAI). Schwartz conducted a study over 20 years ago just as CAI was becoming a vogue form of instruction to attempt to determine whether computer games provided more benefits than traditional forms of phonics instruction. That Schwartz study was relatively small in scope compared to others mentioned with 24 elementary school aged children being divided into two equal groups with one being the experimental and the other the control. The study lasted for 10 weeks and featured students either receiving treatment via the DISTAR Reading program in the Control group or a training session using a computer program that was designed to be similar to an actual video game in its reward system. For example, children engaged in a variety of games such as letter match, word match, and speed reading which featured the use of a microphone and the better they performed on the game the more points they accumulated and the higher level they attained in the game. The GAP Reading Comprehension Tests, Peabody Picture Vocabulary Test, and Clinical Evaluations of Language Functions where given as post and pre tests to measure students abilities. In the post test, only students who performed low on the initial test in the experimental group had significant increases in performance as compared to students in the control group.

            Many students in regards to CAI have interventions that only last a short duration and have small sample sizes of participants. The Hook, Macaruso, and McCabe study of the Lexia CAI program which is phonics intensive had an intervention program that lasted 6 entire months of the school year and used 10 classes total in the study with 5 classes being in the experimental group and 5 being in the control. Another aspect of this study which differentiates it from others is the Lexia computer program was installed directly into classroom computers and was used by all the students in the classroom as part of daily classroom instruction and routines. This aspect of the study may lend it some more practical purpose since the computer program is integrated as a normal part of the classroom as opposed to being done during a special time such as summer school or by “pulling out” students out of normal classroom routines. Students in the control classroom received normal class instruction. It should be noted, that these classrooms where located in an urban school district (Boston) and came from five different Elementary Schools with one control group and one experimental group being housed in each Elementary School. All students were given the The Gates Macginitie Reading Test as post and pre tests in order to gauge performance. All students in the experimental and control groups showed gains on post vs. pre tests but only the Title 1 students showed significant improvements on the GMRT compared to Title 1 students in the control group. (Hook, PE, Macaruso, P, & McCabe, Robert. 2006)

            While the previous study reviewed lasted for an entire year, the post test where only administered once and thus it is inconclusive whether or not any benefits gained by the CAI programs where short of long term. A study by Segers, E, & Verhoeven took a longitudinal look at a similar phonics based computer program effects; conversely, the study is the only one in this review that was not conducted in the United States, but in this case on Dutch Kindergartners in the Netherlands, which may limit its potential for its implication in the United States, but could give us insight into a more diverse set of learners which something that will be discussed in the conclusions and implications for reading development section of this review.

            Segers and Verhoeven specifically sought to determine where there any direct learning games for kindergartners after a year of formal reading instruction assisted by computer programs and whether or not the benefits where similar for native and non-native Dutch speakers. In the experimental group, 5 different computer games where used for a 40 week period. For the first 30 weeks each child used the computer game for 15 minutes a week. For the last 10 weeks children used the games 3 times per week. The fact that 5 different computer programs where used make this study less specific regarding one program in particular and might be able to make more broad claims regarding the efficacy of computer programs in general. Test were given pre-intervention, during Christmas break in the middle of the intervention, at the end of the school year, and 4 months into first grade that measured  the students abilities in  rhyming tasks, phonemic segmentation tasks, auditory blending, and grapheme identification.  There were no major differences between experimental groups and control groups when it came to auditory blending and phonemic segmentation. However, there was significant gains when it came to grapheme knowledge and rhyme. In the case of rhyming tasks, the gains for the immigrant population where so large that they nearly matched native students performance. (Segers, E, & Verhoeven, L. 2005)

            Getting back to studies conducted in the states, Macaruso and Walker conducted a study assessing the efficacy of CAI in advancing literacy skills in kindergarten children. The study was conducted in 6 kindergarten classrooms located in 3 elementary schools in urban school districts outside of Boston with the goal of determining if CAI instruction was beneficial for kindergarteners and whether or not poor performers improved at a different rate than normal performers. The design of the study was focused on eliminating variables that might result in potential invalidity of the results. To achieve this goal, 3 teachers who taught different kindergarten classrooms in the afternoon and morning where used in the study. One of their classes was randomly assigned as the control and the other as the experimental group. Therefore, the students in the control and the experimental groups received the same curriculum, in the same class, and most importantly from the same teacher. The DIBELS Reading Test was used as a pre and post test to assess the students and Gates-Macginitie reading test was used as a post test measure. In the results, significant gains were seen in the experimental group vs the control group when it came to phonological awareness tasks on the post tests. The gains on the post test were even more significant for students who scored lower on the pretests in the experimental group. However, there were no significant advantages for the experimental group on the post test vs. the control group when it came to letter identification or literacy concepts. (Macaruso, Paul, & Walker, Adelaide. 2008).

          Finally, we will review a study conducted by Chambers , Abrami, Cheung, Gifford, R, and Slavin that looks at computer based technology in 2 high poverty and high minority success for all schools. The study focused on the effectiveness of combination CAI with video based instruction as part of the classroom; specifically Allie Ally tutoring games and Reading Reels videos which are short videos usually featuring a puppet character that models reading skills such as blending and segmentation. One school was located in Las Angeles and the other in Las Vegas. Unlike all the prior students, a vast majority of the students in these schools where ELLs with 90% being ELL at the LA school and 70% of the students in Las Vegas. In both cases the nearly all the ELLs came from Spanish speaking households. Students where pre and post tested using the Woodcock Letter-Word identification scale. In addition, they were tested using Woodcock Word Attack, Gray Oral Reading-Test Fluency, Gray Oral Reading Comprehension. The study lasted for the entire school year and in the experimental group all children were instructed using both the Alphie’s Ally CIA program for 20 minutes during tutoring session and using 90 minutes of success for all reading lessons that contained 5 short embedded video clips into the lesson. Post test scores where adjusted for pre-test and the results showed that experimental students scored significantly higher than controls on all 4 of the post tests. (Chambers , B, Abrami, PC, Cheung, A, Gifford, R, & Slavin, RE. 2008).

Conclusions From The Research Review and Implications For Understanding Reading Development

            Many consistent themes appear when looking at all the research. 5 of the studies showed the greatest improvements for children in the experimental groups who scored the lowest on pre-tests. While there is a possibility that all of these similar outcomes did not result because of some inherent benefit of CAI, but rather due to some other factor like regression to the mean, it seems reasonable to speculate that children with the greatest difficulties gain the most because of CAI. These improvements for the children with the greatest deficits were also seen in a variety of different types of studies and even more importantly populations of students ranging from rural schools (Troia , GA, & Whitney , SD. 2003 ), urban schools (Hook, PE, Macaruso, P, & McCabe, Robert.  2006), and even completely different countries such as The Netherlands  (Segers, E, & Verhoeven, L. 2005).

          Unfortunately, another trend that can be garnered by looking at all the studies in the review as a whole is the inconclusiveness of the current research.  Many studies found significant gains in one area of development but not in others. Some studies found little improvement on tasks like blending and segmentation, but significant improvements on other tasks like rhyme and grapheme knowledge.  (Segers, E, & Verhoeven, L. 2005). Furthermore, there are serious questions that whether or not any improvements made would have been greater if the children had undergone some other intensive instructional program such as LiPS which was shown to be more effective in some areas than CAI programs like Fast ForWord and Earobics. (Pokorni, JL, Worthington, CK, & Jamison, PJ. 2004)

                        The length of the CAI intervention also seemed to play a part in its effectiveness. Interventions that lasted only weeks to a month did not seem to end up having as many long term gains as longer CAI interventions. This would seem to suggest what may already feel like intuitive knowledge that the longer one spends learning a task or piece of information the more adept they will be at the task. It was unfortunate that most of the studies regarding CAIs were short term instead of longitudinal so we could not determine if CAIs can help reverse aforementioned Mathew Effects in Reading over time.

            Another conclusion that can be drawn from the review is to take any claims made by creators of CAI programs with a grain of salt. Fast ForWord’s creators and backers conducted studies whose results were not usually verified when the claims were examined by other researchers. This is something that school districts and teachers need to be aware of before they commit to spending large sums of money on a program such as Fast ForWord.

Application In Research In Practice

            There are many aspects of CAIs that appear ripe for further study in the future. There does not appear to be much research on the effectiveness of CAI for ELLs. Additionally, none of the CAI programs in this review contained a bi-lingual component. With an ever increasing number of ELLs in especially native Spanish speakers across the country, research into CAI programs that are specifically designed for  ELL students seems necessary. If programs that incorporate Spanish into the actual instruction do not exist at the moment, then research should focus on how to create effective programs.

            More research should also be done regarding the cost of CAI programs vs. other intervention programs which use actual humans as tutors. In many of the studies such as the comparison of Fast ForWord, Earobics, and LiPS, similar benefits were seen with CAI and human tutors. ( JL, Worthington, CK, & Jamison, PJ. 2004) A meta-study that analyzes the cost of a CAI per student vs. the cost of tutors for a school system would go a long way to help determine what is the best use of a school’s limited resources to benefit the most number of children.

Conclusions

          There is nothing to suggest that CAI can replace traditional classroom instruction in terms of effectiveness. That being said, in every study with CAI experimental groups vs. control groups that did not receive any intervention, the CAI group did better or at the very least equal on post test measures. To put it another way, there where no cases of CAI adversely affecting a child performance. Hence, it would behoove many school systems to try to find a way for students especially those on the lower achievement rung to have access to CAI in some way. The ABRACADABRA model which is a free online program is an interesting one. Because anyone can access it with internet access, it stands to reason that it can be easily used in conjunction with traditional in class instruction if the students have access to the internet at home or from community or school library. That leads us to our final conclusion that greater access to CAI technology could only benefit reading instruction as a whole and as a society we should look for ways to make that access as cheap and easy as possible.

References

(2008). ABRA description . Retrieved from http://grover.concordia.ca/abra/current/pd/parent/pages/what_is_abra.php

Chambers , B, Abrami, PC, Cheung, A, Gifford, R, & Slavin, RE. (2008). Technology infusion in success for all: reading outcomes for first graders. The Elementary School Journal , 109(1), 1-14.

Gillman, RS. (1999). Computer-assisted language intervention using fast forword. Language, Speech, and Hearing Services In Schools, 30, 363-370.

Hook, PE, Marcaruso, P, & Jones, S. (2001). Efficacy of fast forword training on facilitating acquisition of reading skills by children with reading difficulties-a longitudinal study . Annals of Dyslexia , 51, 75-96.

Hook, PE, Macaruso, P, & McCabe, Robert. (2006). The Efficacy of computer-based supplementary phonics for advancing reading skills in at-risk elementary students . Research In Reading , 29(2), 162-172.

Macaruso, Paul, & Walker, Adelaide. (2008). The Efficacy of computer assisted instruction for advancing literacy skills in kindergarten children. Reading Psychology, 29, 266-287

(2009). Metiri group fastforward. Retrieved from http://www.metiri.com/kentucky2001/fastforward.htm

Pokorni, JL, Worthington, CK, & Jamison, PJ. (2004) Phonological awareness intervention: comparison of fast for word, earobics, and lips. The Journal Of Educational Research, 97(3), 147-156.

Savage, RS, Deault, L, & Abrami, Ph. (2009). A Randomized controlled trial study of the abracadabra reading intervention program in grade 1. Journal of Educational Psychology , 101(3), 590-604.

Segers, E, & Verhoeven, L. (2005). Long-term effects of computer training of phonological awareness in kindergarten. Journal of Computer Assisted Learning, 21, 17-27.

Schwartz, S. (1988). A Comparison of componential and tradition approaches to training reading skills.Applied Cognitive Psychology , 2, 189-201.

Stanovich, K. (1986). Matthew effects in reading: Some consequences of individual differences in the acquisition of literacy. Reading Research Quarterly, 21(4), 360-407.

Troia , GA, & Whitney , SD. (2003 ). A Close look at the efficacy of fast forword language for children with academic weakness . Contemporary Educational Psychology , 28(3), 466-493 .