Study of Psychoeducation and Reading for Kids (SPARK) (SPARK)

The proposed project is to evaluate the effectiveness of a psychoeducational tutoring intervention, first instituted in the 1990s, refined since that time, and delivered since 2014 by a nonprofit, the Organization for Psychoeducational Tutoring (OPT). The research, Study of Psychoeducation and Reading for Kids, (SPARK) is funded by gifts to the Upstate Foundation. The research, conducted under auspices of Upstate Medical University, is a collaborative project with OPT to study the efficacy of the methods OPT is using.

The study has one primary aim: To evaluate the effectiveness of the intervention on two primary outcomes: reading skill and overall psychological functioning.

The study has one secondary aim: To assess the feasibility and validity of some novel performance measures of psychological functioning.

For the reading intervention, the major outcome variable is the rate or student advancement, in grade or age equivalents gained per year, during tutoring. This rate will be as contrasted with the rate established for the normative samples for the tests, i.e. one grade or age equivalent per year. Thus the normative samples function as a comparison group. The rate will also be contrasted with the child's rate of progress in reading before tutoring. The latter is the comparison reported in Strayhorn (2005) (where rates before and during tutoring were 0.44 and 2.0 grade equivalents per year, respectively). The Slosson Oral Reading Test, a measure of reading recognition, and the text comprehension subscale of the Test of Reading Comprehension will be our major reading measures.

For psychological functioning, we will analyze pre-post differences in parent and teacher ratings with several rating scales, including the Strengths and Difficulties Questionnaire, and blind ratings of overall functioning and overall improvement drawn from interviews of parents, teachers, and children. For the Strengths and Difficulties Questionnaire, we will also use the method of "virtual controls" (Strayhorn, 2021, Ford et al., 2009). The method of virtual controls is to find, or construct, regression equations that will predict postscores from prescores and other covariates, in untreated samples. The predicted postscore for each subject, generated by using the regression equation for the particular preintervention numbers for each participant, is the virtual control with which the actual treated postscore is compared. Virtual control methods will strengthen our ability to estimate treatment effects despite the non-randomized design.

The proposed intervention targets a broad spectrum of psychological skills, that are felt to be the opposites of the behavior patterns described for psychiatric disorders. The U.S. National Comorbidity study (Merikangas et al., 2010) provided some evidence for broad-spectrum interventions, as contrasted with those that target only one disorder: 40% of participants with a disorder had a disorder in another class (i.e., anxiety, mood, behavior, substance, and other); 18% of youth with a disorder had disorders in three or more different classes. The strong tendency of psychiatric disorders to be comorbid with one another is a reason for teaching a wide swath of psychological skills rather than putting primary emphasis upon pinpointing diagnoses.

Kazdin and Blase (2011) proposed a "rebooting" of mental health research and practice to reduce the burden of mental illness in the population, arguing that the traditional psychotherapeutic service delivery model failed to bring about sufficient accessibility of evidence-based procedures to have a large enough impact on the mental health of populations. In response to this article, Yates (2011) called for "methods that use less therapist time, less client time, minimize client transportation costs as well as brick-and-mortar space, and use less of other increasingly scarce and costly resources." The proposed intervention represents a "rebooting" by using nonprofessional tutors, no brick-and-mortar space, and surmounting geographical barriers by phone intervention.

The relation between reading difficulties and behavior disorders is strong. Stanton et al., (1990) concluded that the well-known association between low IQ and classroom behavior problems disappears when reading skill is statistically controlled. Morgan et al. (2008) reviewed more than 125 studies regarding the question, "Are reading and behavior problems risk factors for each other?" The answer seems conclusively to be: Yes, there is evidence for causation in both directions. Reading difficulties are prevalent. According to 2022-2023 test results, only 45% of New York public school third graders scored in the "proficient" level in English language arts, and only 30% of students from low income backgrounds did so (Education Trust, 2024).

Children upon entry into the study will be classified as having needs in reading, psychological skills, or both. The outcome data for reading and psychological skills will be analyzed separately, using the subsamples with needs in the area in question.

For those on the reading track, the activities are those described in Strayhorn, (2009), Manual for Tutors and Teachers of Reading. The reading intervention is meant to give the child success experiences in whatever point on the hierarchy of reading challenges is best for that child at any given moment; the hierarchy includes foundation skills of phonemic awareness, spatial awareness, and letter-sound correspondence, and more advanced skills of sounding and blending progressively more complex words and reading progressively more complex text. As soon as the learner is capable of reading text, the content of the reading is meant to teach about psychological skills.

For those on the psychological skills track, the three major activities are: alternate reading from one of the manuals, doing psychological skills exercises, and engaging in unstructured chatting, wherein the tutor attempts to model good listening techniques. The family is sent (at no cost) books that they will use in the intervention; the tutor uses electronic copies of these. The child and the tutor have half hour sessions that are scheduled for 5 or 6 days a week. Several manuals have been created for this project, to be used in alternate reading; the full texts of these are available on the Internet at https://www.josephstrayhornmd.com and at https://optskills.org/manuals.html.

The intervention continues for at least one year for each child. After that, the child and family may elect to continue with the intervention. Throughout the intervention, the tutors and their supervisor continually endeavor to update the content of the sessions so as to be appropriate for the child's current level of skill, interest, and need. The goal is that all students eventually reach the psychological skills track as they move from "learn to read" to "read to learn" emphasis. Alternate reading, aloud, and answering comprehension probes, continue to be very effective methods for improving reading skill, even after the child has ceased the official reading instruction.

Parent training is an important component of any psychological-skills-oriented intervention for children. The plan for this project is to use a "questionnaire that teaches." This presents 8 of the big ideas communicated by parent training, and asks parents how much they use them and how much they would like to use them more often. At intervals of every two months, we will ask parents these questions in one-to-one phone calls conducted by one of the OPT staff. We will find out what the results have been for trying any of these ideas so far. We will also let any interested parents know that 2 books written by Strayhorn on parenting are available in full text on the Internet (Strayhorn 2015, 2018).

Investigators will, when appropriate, remind families of alternatives to participation, in addition to the reminder in the consent form. There are two major societal institutions pursuing the same goals of the intervention: schools and clinical mental health services. Families will be informed that their participation in the intervention does not rule out their obtaining any additional services from schools, mental health service providers, or anyone else; we ask that they report additional services children receive.

Questionnaires rating children's behavior can be inaccurate. Accordingly, there is one more primary outcome variable: blind ratings of the child's global functioning based on write-ups of "evaluations" of the child's psychiatric and academic status. Interviews of parents, and possibly teachers and students themselves, will seek to give as complete a picture of the child's functioning as possible. Questions delve into levels of the psychological skills forming the outline of the psychoeducational curriculum. The written narratives which compile the data will then be rated by blind raters using a scale of overall functioning.

During the course of the intervention, we will be administering several performance measures of psychological skills, for development during the study. These seek to measure mental health functioning by testing rather than relying solely on ratings. The promising test procedures will be studied for internal reliability (using coefficient alpha) and test-retest reliability, and for validity using correlations with other measures.

The project poses minimal risk. This conclusion is bolstered by the fact that the intervention has been carried out continuously by phone for the last 27 years, and before that a similar intervention was carried out in person at a school. There have been no complaints of harm done.

Possible benefits include making gains in reading, learning anger control skills, reducing anxiety, being more able to feel good about one's positive examples of psychological skills, learning to have better relationships with people, learning to improve the decision-making process, learning to become more psychological-minded and to increase the vocabulary with which one thinks about relationships and emotions.

For the group of children where reading is targeted, we will calculate the rate of progress before tutoring by dividing the grade level at which the child reads by the number of years of education so far, starting with the kindergarten year. The Slosson Oral Reading Test, 3rd Edition (Slosson & Nicholson, 2002), a measure of reading recognition, will be the primary measure of reading competence. The Test of Reading Comprehension, 4th edition, text comprehension subtest, (Brown et al., 2009) will be another important measure of reading competence. For both, to calculate the rate of progress during tutoring, we subtract the grade level at the beginning from that at the measurement point, and divide by the number of years of tutoring so far. To compute prior to tutoring, we take the child's grade equivalent at the beginning of tutoring and divide by the number of years of schooling, starting with kindergarten. We compare the two rates with a t-test, and a 95% confidence interval will be obtained. We also will compare the rate of progress during tutoring to that of the norming sample, i.e., 1 grade equivalent per year.

For the group of children who started with scores low enough to be in the psychological skills group, the method of virtual controls (Strayhorn, 2021) will be used. Ford et al. (2009) worked out a regression equation predicting postscores without intervention on the Strengths and Difficulties Questionnaire (SDQ) (Goodman, 2001), using predictors from the preintervention questionnaire. During the course of the intervention we will check this regression equation using time 1 and time 2 SDQ scores from several other longitudinal studies. The scores predicted by regression will then be compared with the actual postintervention scores with a t-test, and a 95% confidence interval for the difference will be obtained. The virtual controls strategy avoids some of the ethical and practical problems of randomized clinical trials.

To obtain measures of global functioning for children in the psychological skills track, blind raters will read the written assessment summaries and rate the child's functioning at pre and post using the Children's Global Assessment Scale. For this measure, means and standard deviations of functioning ratings will be computed for pre and post score and confidence intervals computed for improvement scores.