Promoting Interest and Performance in High School Science Classes

Abstract

We tested whether classroom activities that encourage students to connect course materials to their lives will increase student
motivation and learning. We hypothesized that this effect will be stronger for students who have low expectations of success.
In a randomized field experiment with high school students, we found that a relevance intervention, which encouraged students
to make connections between their lives and what they were learning in their science courses, increased interest in science
and course grades for students with low success expectations. The results have implications for the development of science
curricula and theories of motivation.

Full text

Hulleman Relevance in Science 1




Promoting Interest and Performance in High School Science Classes
Chris S. Hulleman1* and Judith M. Harackiewicz2
1Department of Graduate Psychology, James Madison University
2Department of Psychology, University of Wisconsin-Madison
*Corresponding author. E-mail: hullemcs@jmu.edu
WORD COUNT: 2372
In press, Science, October 5, 2009.

Hulleman Relevance in Science 2

Summary
A randomized field experiment demonstrated that a motivational intervention designed to help
high school students make connections between course material and their lives increased interest
and performance in high school science courses.
Abstract
We test whether curriculum that is relevant and personally meaningful will increase student
motivation and learning. We hypothesize that this effect will be stronger for students who have
low expectations of success. In a randomized field experiment with high school students, we find
that a relevance intervention, which encouraged students to make connections between their lives
and what they were learning in their science courses, increased interest in science and course
grades for students with low success expectations. The results have implications for the
development of science curricula.
Text
Many educators and funding agencies share the belief that making education relevant to
students’ lives will increase engagement and learning (1-6). However, little empirical evidence
supports the specific role of relevance in promoting optimal educational outcomes, and most
evidence that does exist is anecdotal or correlational (1-3, 5, 7, 8). The purpose of our research is
to demonstrate how an intervention specifically designed to enhance the relevance of science to
students’ lives can enhance interest in science and classroom performance, particularly for
students who are most at risk for being disengaged from school.
Numerous curricular reform efforts have emphasized applying science to students’ lives, such as
providing out-of-school research experiences (7, 9), creating learning modules for specific topics
(e.g., “Acids, Bases, and Cocaine Addicts,” 10, 11), developing an undergraduate course (e.g.,
“The Biology and Chemistry of Everyday Life”, 12-13), and redesigning the academic structure
of entire high schools (14). For example, the Metro Nashville Public School District redesigned
several of its high schools into career academies within which students can choose from
thematically-focused learning communities (15). The intention of these career academies is to
enable students to “connect what they learn in school with their career aspirations and goals”
(14). Although many of these programs have produced positive outcomes, such as improvements
in retention in academic programs (13) or performance on achievement tests (7, 10-12), it is not
clear that these effects were due to personal relevance. These educational reforms are multi-
faceted, and an emphasis on relevance is just one of several components that may have
contributed to the programs’ outcomes. For example, other potentially effective components are
small group instruction (12, 16), repeated exposure to the material (10), individual mentoring and
teaching (14), individualized and/or team-based projects (7, 13, 16), hands-on activities (9),
visualization exercises (11), increased autonomy (6), and increased knowledge development (17,
18).

Hulleman Relevance in Science 3

Programs that emphasize personal relevance may be particularly empowering for students who
are disengaged from school due to a lack of confidence. Students can become energized if they
believe they are competent in science and can successfully perform classroom tasks. As
described by expectancy-value models of motivation (19), both an individual’s expectancy for
success and their perception of value for the activity facilitate student motivation. Research on
expectancies reveals that expecting to successfully perform a task leads to greater persistence,
performance, and interest in academic activities (19-20). Thus, students who do not believe that
they can do well in the classroom are at risk for performing poorly and becoming less interested
in academics.
In addition to lacking confidence, students with low success expectancies may not perceive, or
may have a harder time perceiving, relevance and value in their schoolwork (21). These students
may require external support – from teachers or classroom activities – to foster or maintain task
engagement (22). Interventions that facilitate the perception of relevance in a topic might
promote attention and learning for students with low success expectancies (23). Instead of
withdrawing from the activity, these students may become energized as they discover reasons for
exerting effort and becoming more involved in learning (24). In contrast, more confident
students may not need this type of motivational boost because their effort and involvement in
school are already strong (22, 24).
Reduced interest in academics is particularly problematic for long-term outcomes such as
educational and career choices. Research on the development of interest (i.e., experiencing
positive affect, value, and knowledge with an activity) demonstrates that interest is a more
powerful predictor of future choices than prior achievement or demographic variables (22-25).
For example, Harackiewicz and colleagues (25) found that interest in an introductory
undergraduate psychology course during freshman year was more predictive of subsequent
course taking and majoring in psychology over a seven-year span than were grades from that
introductory course. Interest development can begin in situations that promote student
engagement with the material (i.e., situational interest; 24). If students repeatedly experience
situational interest in relation to a particular topic, they may eventually develop a more enduring
interest in the topic (i.e., individual interest). A crucial factor in the progression from situational
to individual interest is finding personal meaning and relevance in a topic (24). Perceiving a
topic to be useful and relevant for other activities or life goals (i.e., utility value) predicts both
subsequent interest and performance (8).
Making science courses personally relevant and meaningful may engage students in the learning
process, enable them to identify with future science careers, foster the development of interest,
and promote science-related academic choices (e.g., course enrollment, pursuit of advanced
degrees) and career paths. The first step in this process is to investigate whether emphasizing
relevance in the classroom promotes interest and performance in science courses. Thus, the goal
of our research was to examine the effectiveness of a curricular intervention on interest and
performance in high school science classes, particularly for students with low performance
expectations.

Hulleman Relevance in Science 4

We conducted a randomized field experiment of a motivational intervention that was designed to
help students make connections between their high school science classes and their lives. The
intervention was embedded within the entire semester of ninth-grade science courses. We
investigated whether this intervention would increase student interest in science, performance in
the course, and interest in science-related careers compared to a control condition where students
wrote summaries of the material they were studying. Because students wrote about science topics
in both conditions, knowledge activation was controlled, and the conditions differed only in
terms of personal relevance activation. We predicted that the relevance intervention would
promote interest in science and performance in the course, particularly for students with low
performance expectations. Subsequently, we expected that increased science interest would lead
to more interest in science-related courses and careers.
Participants were 262 high school students taught by seven science teachers (biology, integrated
science, physical science) from two high schools in a small, mid-western city in the US. Students
were 92% ninth-graders (8% tenth-graders), 52% female, 66% Caucasian, 15% African-
American, 12% Asian, and 8% Hispanic. The analysis covered one academic semester. Students
were randomly assigned within each classroom to either write about the usefulness and utility
value of the course material in their own lives (relevance condition, N = 136), or write a
summary of the material they were studying (control condition, N = 126). Teachers were
informed that the research concerned the effectiveness of writing assignments, but were blind to
our hypotheses and students’ experimental conditions. To ensure this, the researchers randomly
assigned students to conditions at the beginning of the semester by giving students booklets with
identical covers, but with different instructions inside depending upon experimental condition.
Students completed their essays in these books, which were collected by the researchers after
each assignment, every three or four weeks starting at the beginning of the semester. Students
wrote from 1 to 8 essays (M = 4.7, SD = 1.4) throughout the semester. The researchers provided
teachers with information regarding whether students had completed the essays, but teachers
remained blinded to condition throughout the semester (26).
Students’ success expectancies and initial interest in science were measured at the beginning of
the semester. Student’s interest in science and future plans for science-related courses and
careers were measured at the end of the semester (see Table S2 for self-report items; 26). Second
quarter grades were obtained from school records for one of the high schools (N = 100 students).
The data were analyzed using multiple regression with dummy codes representing the nesting of
students, teachers, and schools (26). The focal predictor was the interaction between the dummy
code for experimental condition (0 = control, 1 = relevance) and student’s performance
expectations for the course. We predicted that this interaction term would be negative, such that
the intervention effect would be more positive for those with low as opposed to high
performance expectations. To examine effects of sex and race, and whether the condition effects
varied by sex and/or race, the regression models included dummy codes for sex, race, and their
interactions with condition (26).
As predicted, there was a significant negative interaction between the relevance intervention and
students’ expectations for success on science interest (β = -0.11, p = .05), and the same negative

Hulleman Relevance in Science 5

interaction was also significant on second quarter grades (β = -0.18, p = .03). The predicted
values from the regression equation indicate that students with low success expectancies (one
standard deviation below the mean) reported more interest in science at the end of the semester
(and received higher course grades) in the relevance condition than in the control condition,
whereas students with high success expectancies (one standard deviation above the mean)
reported similar levels of interest (and course grades) regardless of experimental condition (Fig.
1). There were no statistically significant interactions with gender or race, indicating that
intervention did not have differential effects (Tables S4 and S5). Finally, interest in science at the
end of the semester was a significant predictor of interest in future science-related courses and
careers (β = .58, p < .01; see Table S6).
Our results demonstrate that encouraging students to make connections between science course
material and their lives promoted both interest and performance for students with low success
expectancies. The effect on performance was particularly striking, as students with low success
expectancies improved nearly two-thirds of a letter-grade in the relevance condition compared to
the control condition, which is comparable to other social-psychological interventions aimed at
reducing the black-white achievement gap (27). Differences between conditions for students with
high success expectancies were not statistically significant. These results provide experimental
support for expectancy-value models of motivation that hypothesize that perceived value is an
important contributor to interest, performance, and future plans (19-21).
Although our experimental design included randomizing students within classrooms, evaluating
the effects of the intervention over time, and assessing change in our dependent variables by
including pre- and post- measures in the analyses, this single study requires replication before
generalizations can be made about more diverse settings and students. In addition, although the
control condition was designed to activate students’ content knowledge, further investigations of
the cognitive processes instigated by the intervention are warranted. Nonetheless, our results
show how motivational principles can be utilized to increase interest and performance in science
courses early in high school. Our intervention was effective in raising interest and performance,
was easy to implement, and required few external resources. This type of motivational
intervention can be easily incorporated into almost any course with little cost to the instructor.

Hulleman Relevance in Science 6

Figure Legend
Figure 1. Science interest and course grades as predicted by the relevance intervention and
performance expectations. Predicted values are computed from the multiple regression equation
for the interaction between the relevance intervention and performance expectations (Low = -1
SD, High = +1 SD) on final course interest and second quarter grades. Error bars represent +/- 2
SEM (0.12 for interest and 0.28 for grades).
0
1
2
3
4
LowExpectations HighExpectations
ScienceInterest
Control Relevance
0.0
1.0
2.0
3.0
4.0
LowExpectations HighExpectations
SecondQuarterGrade
Control Relevance

Hulleman Relevance in Science 7

References
1. M. R. Lepper, J. Henderlong, in Intrinsic and extrinsic motivation: The search for
optimal motivation and performance, C. Sansone, J. M. Harackiewicz, Eds. (Academic
Press, Inc., San Diego, 2000), pp. 257-307.
2. National Research Council, How people learn: Brain, mind, experience, and school.
National Academies Press, Washington, DC (2000).
3. National Research Council, Engaging schools: Fostering high school students’
motivation to learn. National Academies Press, Washington, DC (2004).
4. P. R. Pintrich, D. H. Schunk, Motivation in education: Theory, research, and
applications (2nd Ed.). (Prentice Hall, Upper Saddle River, NJ, 2002).
5. T. J. Newby, J Educ Psy 83, 195-200 (1991).
6. K. Mitchell et al., Rigor, relevance and results: The quality of teacher assignments and
student work in new and conventional high schools. (American Institutes for Research
and SRI International, Washington, DC, 2005).
7. D. I. Hanauer, et al., Science 314, 1880-1881 (2006).
8. C. S. Hulleman et al., J Educ Psy 100, 398-416 (2008).
9. S. A. Tucker, D. L. Hanuscin, C. J. Bearnes, Science 319, 1621-1622 (2008).
10. N. C. Kwiek, M. J. Halpin, J. P. Reiter, L. A. Hoeffler, R. D. Schwartz-Bloom, Science
317, 1871-1872 (2006).
11. M. C. Linn, H. Lee, R. Tinker, F. Husic, J. L. Chiu, Science 313, 1049-1050 (2006).
12. L. A. Denofrio, B. Russell, D. Lopatto, Y. Lu, Science 318, 1872-1873 (2007).

Hulleman Relevance in Science 8

13. N. L. Fortenberry, J. F. Sullivan, P. N. Jordan, D. W. Knight, Science 317, 1175-1176
(2007).
14. S. H. Russell, M. P. Hancock, J. McCullough, Science 316, 548-549 (2007).
15. “Career academy,” (Helping America’s Youth Web site, 2008,
http://guide.helpingamericasyouth.gov/programdetail.cfm?id=96).
16. “Metro Nashville Public School career/thematic centers for 2008-09,” (Metro Nashville
Public Schools Web site, http://www.hillwoodhs.mnps.org/page35350.aspx).
17. F. M. Newmann, H. M. Marks, A. Gamoran, Am J Educ 104, 280-312 (1996).
18. F. Newmann, A. S. Bryk, J. K. Nagaoka, Authentic intellectual work and standardized
tests: Conflict or coexistence? (Consortium on Chicago School Research, Chicago,
2001).
19. J. Eccles et al., in Achievement and achievement motives: Psychological and sociological
approaches, J. T. Spence, Ed. (Freeman, San Francisco, 1983), pp. 75-146.
20. K. A. Updegraff, J. S. Eccles, B. L. Barber, K. M. O’Brien, Learn Ind Diff 8, 239-259
(1996).
21. J. E. Jacobs, S. Lanza, D. W. Osgood, J. S. Eccles, A. Wigfield, Child Dev 73, 509
(2002).
22. S. Hidi, J. M. Harackiewicz, Rev Educ Res 70, 151-179 (2000).
23. S. Hidi, Rev Educ Res 60, 549-571 (1990).
24. S. Hidi, K. A. Renninger, Educ Psy 41, 111-127 (2006).
25. J. M. Harackiewicz et al., J Educ Psy 94, 562–575 (2002).
26. See Supplementary Online Material.
27. G. L. Cohen et al., Science 313, 1307-1310 (2006).

Hulleman Relevance in Science 9

28. Acknowledgments: This article is based on a doctoral dissertation submitted by Chris S.
Hulleman to the University of Wisconsin-Madison under the supervision of Judith M.
Harackiewicz. Thanks are extended to Martha Alibali, Geoffrey Borman, Patricia Devine,
and Adam Gamoran for their service on the dissertation committee; to the Madison
Metropolitan School District for their assistance with data collection; and to Janet Hyde,
Steven McFadyen-Ketchum, Clifford Thurber, and Teresa Hulleman for their comments
on earlier versions of this manuscript. This research was supported in part by grants from
the Department of Psychology and the Institute for Education Sciences, U.S. Department
of Education, through Award #R305C050055 to the University of Wisconsin-Madison
and Award # R305B050029 to Vanderbilt University.

Hulleman Relevance in Science 10

Supporting Online Material
Promoting Interest and Performance in High School Science Classes
Chris S. Hulleman*, Judith M. Harackiewicz
*Corresponding author. E-mail: hullemcs@jmu.edu
This supplement contains:
Methods and Results
Tables S1 to S8
Appendix: Intervention Instructions
References

Hulleman Relevance in Science 11

Promoting Interest and Performance in High School Science Classes
Methods and Results
Participants. The overall sample consisted of 262 students from two different high
schools within the same school district in a small, mid-western city. Students were nested within
seven teachers and 19 classrooms. Eight of the classes were required freshman biology; nine
classes were part of an integrated science program, primarily for ninth-graders, that covered
biology, chemistry, physics, and geology; and two classes were elective physical science courses.
Participants were 92% ninth-graders (8% tenth-graders), 52% female, 66% Caucasian, 15%
African-American, 12% Asian, and 8% Hispanic. This was the overall sample used in the
analyses in interest. For the analyses predicting course grades, we were able to obtain course
grades for a sub-sample of these students (N = 100). In the grade sub-sample, students were 54%
female, 68% Caucasian, 23% African-American, 4% Asian, and 5% Hispanic.
Procedure.During the second week of the semester, students were randomly assigned to
one of two writing conditions – in the relevance condition, students were asked to apply what
they were learning in class to real life, and in the control condition students were asked to write a
summary of what they were learning (see Appendix for complete instructions). In both
conditions, students were asked to select a topic that was currently being covered in class (e.g.,
photosynthesis). Following the selection of a topic, students in the relevance condition (N = 136)
were asked to write a one-paragraph essay that applied the topic to their life or to the life of
someone they knew. Students in the control condition (N = 126) were asked to write a one-
paragraph summary of the topic they selected. The writing intervention was part of the course
syllabus and was completed for course credit. In order to keep teachers blind to each student’s
experimental condition, the instructions for the writing assignment were placed inside an essay

Hulleman Relevance in Science 12

booklet that students received during the second week of the semester. Students placed their
name on the outside of the book and wrote their essays inside each book. Teachers were
instructed to have students complete an essay as part of their review activities one to two days
prior to each test. The teachers were asked not to read or grade the essays. Rather, after each test
the researcher collected the booklets, read the essays, and assigned them a score from 0 to 2
points. The instructor then used this score as a basis to award course credit. The researcher
returned the booklet for the student to use prior to the next test. The number of essays that
students wrote during the semester varied in each classroom according to how many tests were
offered by the teacher (M = 4.7, SD = 1.4; see Table S1).
Manipulation check. Two trained research assistants, blind to the hypotheses of the study,
read each of the essays and rated the extent to which students made connections to their lives
(88% inter-rater reliability, differences resolved through discussion) on a scale from 0 (no
connections), 1 (moderate connections), to 2 (strong connections; see Table S3 for examples of
student writing). The essay ratings indicated that students in the relevance condition made more
connections in their essays (M = 0.91, SD = 0.67) than those in the control condition (M = 0.10,
SD = 0.31; t(258) = 12.37, p < .001). In order to examine whether the essays were of equivalent
quality, the essays were coded for quality of writing on a scale from 0 (no response or
unintelligible), 1 (some structure but unclear, clear but disorganized), to 2 (logical structure,
clear explanation). As expected, the quality of student writing was similar in relevance (M =
1.87, SD = 0.23) and control conditions (M = 1.85, SD = 0.29). Finally, in order to examine the
amount of effort students put into writing the essays, the number of sentences written in each
essay was counted. The average number of sentences written by students was significantly

Hulleman Relevance in Science 13

greater in the control (M = 4.49, SD = 2.29) than the relevance condition (M = 3.92, SD = 1.63, p
=.02, d = 0.28).
Measures. Students’ success expectancies, initial interest in science, perceptions of utility
value of the course material, mastery-approach goals (the goal to learn and improve), and
performance-approach goals (the goal to perform better than others) were measured at the
beginning of the semester. These items were adapted from measures used in previous research
(1, 2). Student’s interest in science and future plans to pursue science-related courses and careers
were also measured at the end of the semester. Table S2 presents the self-report items and scale
reliabilities for all self-report measures. Student’s course absences during the first quarter of the
semester (first nine weeks of the semester) and second quarter grades (second nine weeks of the
semester) were obtained from course records for a sub-sample of students (N = 100).
Analytic approach. The data were analyzed using multiple regression. Dummy codes
were used to represent the experimental conditions (0 = control, 1 = relevance) and the fixed
effects of teachers (3). Six dummy codes were therefore used to represent the seven teachers in
the study (one teacher was chosen at random to be the comparison group). Dummy codes were
also used to represent the effects of sex (0 = male, 1 = female) and race. For the race analyses,
Caucasian students were the comparison group and separate dummy codes were created for
Asians, Blacks, and Hispanics. The exception was for the grade analysis which, due to decreased
sample size, one dummy code represented race (0 = Caucasian, 1 = non-Caucasian). Additional
covariates included performance expectations, initial interest in science, initial perceptions of
utility value of science, and achievement goals. Student attendance was also included as a
covariate when predicting grades. The focal predictor was the interaction between experimental

Hulleman Relevance in Science 14

condition and performance expectations. The interaction term was expected to be negative, such
that the effect of the intervention is expected to be more positive for those with low performance
expectations than those with high expectations.
For the analyses, all continuous main effect terms were standardized, and multiplicative
two- and three-way interaction terms were created with these variables (4). Interactions that were
significant on any measure were retained in all models, but non-significant interactions were
trimmed from models. To interpret significant interaction effects from these analyses, we
computed predicted values (Ŷs) for representative high and low groups (one standard deviation
above and below the mean) from the regression equations using the unstandardized coefficients.
The results from the multiple regression analyses predicting interest, course grades, and future
plans can be found in Tables S4, S5, and S6 respectively.
Results. As predicted, there was a significant negative interaction between the relevance
intervention and students’ expectations for success on science interest (β = -0.11, p = .05), and
the same negative interaction was also significant on second quarter grades (β = -0.18, p = .03).
The predicted values from the regression equation indicate that students with low success
expectancies (one standard deviation below the mean) reported more interest in science at the
end of the semester (and received higher course grades) in the relevance condition (ŶInterest =
3.91, ŶGrades = 3.23) than in the control condition (ŶInterest = 3.55, ŶGrades = 2.43), whereas
students with high success expectancies (one standard deviation above the mean) reported
similar levels of interest (and course grades) in the relevance (ŶInterest = 3.67, ŶGrades = 2.81) and
control conditions (ŶInterest = 3.76, ŶGrades = 3.03). Notably, there were statistically significant
main effects of sex and race for grades, but not interest, suggesting that females received higher

Hulleman Relevance in Science 15

grades (Ŷ = 3.03) than males (Ŷ = 2.43), and Caucasian students received higher course grades
(Ŷ = 3.34) than minority students (Ŷ = 2.12). Total absences was a negative predictor of second
course grades (β = -0.37, p = .01), indicating that students with more absences received lower
grades than students with fewer absences from school. Finally, interest in science at the end of
the semester was a statistically significant predictor of future education and career plans with
science (β = .58, p < .01; see Table S6).
Treatment compliance and dosage. Nearly all of the students who were randomly
assigned to an experimental condition actually received the treatment (260 of 262 students).
Thus, the impact of treatment compliance (“no-shows”, 5) on the treatment effect is negligible.
However, the strength or dosage of the manipulation, in this case the quality of connections
students made between their lives and the material, was variable within the relevance condition
(Table S7). To evaluate the dosage effect within the relevance condition, we compared the mean
levels of final science interest for students who made no connections with the material (M = 3.34,
SD = 0.96) to students who made moderate (M = 3.35, SD = 1.01) and strong connections (M =
3.49, SD = 1.12). The effect size indices indicated that there was no dosage effect between
students in the relevance condition who made no connections and those who made moderate
connections (d = .00), and a small effect between students who made no connections and those
who made strong connections (d = .11).
Predicting attendance. Although we use first quarter attendance as a covariate in the
analyses, it is possible that the relevance intervention might impact attendance. To this end, we
predicted the proportion of days attended during the semester from the regression models used
above. As presented in Table S7, the results indicated that increased first quarter absences were a
negative predictor of days attended during the semester (β = -0.34, p = .02), and that African-

Hulleman Relevance in Science 16

American students were less likely to attend school (β = -0.26, p = .05). Importantly, there were
no statistically significant effects of the intervention, or interactions with the intervention, on
days attended during the semester.

Hulleman Relevance in Science 17

Table S1
Frequency table of number of essays completed
Number of essays
completed Frequency Percent
Cumulative
Percent
1 3 1.1 1.1
2 2 .8 1.9
3 46 17.6 19.5
4 65 24.8 44.3
5 108 41.2 85.5
6 6 2.3 87.8
7 13 5.0 92.7
8 19 7.3 100.0
Total 262 100



Hulleman Relevance in Science 18

Table S2
Self-reported survey items 1
Construct Reliability
(Cronbach’s α)
Items
Expectancies for
Success
.73 I expect to do well in this class.
Considering the difficulty of this course and my
skills, I think I will do well in this class.
Interest .84 I think the field of science is interesting.
To be honest, I just don’t find science interesting.
I think what we’re learning in this class is
interesting.
Utility Value .82 I can apply what we are learning in science class to
real life.
I think what we are studying in science class is
useful to know.
I can see how what I learn from science applies to
life.
Mastery-approach
Goals
.86 My goal in this class is to learn as much as I can.
I want to learn as much as possible in this class.
Performance-approach
Goals
.86 I want to do better than other students in this class.
It is important for me to do well compared to other
students in this class.

Hulleman Relevance in Science 19

Future Plans .84 My experience in this class makes me want to take
more science courses.
I want to have a job that involves science some day.
I plan on taking more science courses even when I
don't have to.
I am not really interested in using science in my
future career.

Hulleman Relevance in Science 20

Table S3
Examples of student writing in the relevance condition
Topic (Course) Writing Sample
The metric system
(Physics)
My family runs a dairy farm. We measure how much milk is in the cooler.
The milkman measures the milk and takes it to the cheese factory and he
had to tell them how much we have and then they will pay us money. It is
important to measure accurately so we need a budget for the future. We
also measure the food for the cows.
The eight
characteristics of
life (Biology)
The reason I chose characteristics of life because when I grow up I want to
be a doctor. Doctors need to know about the characteristics of life in order
to help their patients. This information can help me when I have to take
advanced biology in college. My brothers are always getting sick and
knowing about homeostasis and virus can help me find out why they’re
always getting sick. And I want to be a laboratory technician. They need to
know about DNA and stuff.
Graphing
(Chemistry)
Graphing is important part of life because when you’re trying to compare
different data the graph is the best way to go. For an example, my
grandmother and aunt work at a retirement home and they need to decide
dosages per day, meals, and etc. Graphing out all the data they have will
[help them] come out with a resolution. This applies to college where I
want to go someday.


Hulleman Relevance in Science 21

Table S4
Multiple regression results for science interest
Step 1 Step 2
b SE β p b SE β p
Intercept 3.84 0.14 0.00 3.65 0.18 0.00
Group -0.15 0.10 -0.08 0.15 0.13 0.26 0.07 0.61
Performance Expectations -0.03 0.06 -0.03 0.65 0.10 0.10 0.11 0.28
Sex 0.00 0.10 0.00 0.99 0.24 0.15 0.12 0.12
Asian 0.23 0.17 0.07 0.18 0.28 0.25 0.09 0.26
Black -0.12 0.15 -0.04 0.43 0.02 0.22 0.01 0.92
Hispanic -0.19 0.22 -0.05 0.38 -0.05 0.30 -0.01 0.88
Initial Interest 0.47 0.07 0.48 0.00 0.45 0.07 0.46 0.00
Initial Utility Value 0.05 0.07 0.05 0.42 0.08 0.07 0.09 0.22
Mastery-approach Goals 0.05 0.07 0.05 0.52 0.06 0.07 0.06 0.43
Performance-approach
Goals
0.03 0.06 0.03 0.62 0.03 0.06 0.03 0.61
Teacher Dummy Codes
Teacher 1 -0.06 0.23 -0.02 0.78 0.06 0.30 0.02 0.84
Teacher 2 -0.13 0.19 -0.04 0.52 0.13 0.27 0.04 0.64
Teacher 3 0.09 0.20 0.03 0.67 -0.02 0.27 -0.01 0.95
Teacher 4 -0.56 0.17 -0.22 0.00 -0.57 0.24 -0.22 0.02
Teacher 5 -0.35 0.18 -0.12 0.05 -0.52 0.27 -0.18 0.06
Teacher 6 -0.25 0.17 -0.10 0.13 -0.24 0.24 -0.10 0.31
Interactions with Group
Group * Perf. Exp. -0.22 0.11 -0.11 0.05
Group * Sex -0.39 0.21 -0.10 0.06
Group * Black -0.35 0.30 -0.06 0.25
Group * Asian -0.17 0.35 -0.03 0.63
Group * Hispanic -0.03 0.45 -0.03 0.95
Teacher 1 * Group -0.38 0.46 -0.08 0.41
Teacher 2 * Group -0.45 0.38 -0.10 0.24
Teacher 3 * Group 0.41 0.42 0.08 0.32
Teacher 4 * Group 0.02 0.33 0.01 0.95
Teacher 5 * Group 0.23 0.36 0.07 0.52
Teacher 6 * Group 0.03 0.33 0.01 0.92
R2 .36** .39**
R2 - change .04
Note: b = unstandardized regression coefficient. SE = standard error. β = standardized regression
coefficient.
** p < .01. * p < .05.

Hulleman Relevance in Science 22

Table S5
Multiple regression results for second quarter course grades
Step 1 Step 2
b SE β p b SE β p
Intercept 2.63 0.30 0.00 2.73 0.34 0.00
Group -0.09 0.23 -0.03 0.69 0.29 0.45 0.11 0.51
Performance Expectations 0.00 0.13 0.00 0.99 0.30 0.17 0.04 0.08
Sex 0.67 0.23 0.25 0.00 0.60 0.30 0.22 0.05
White 0.38 0.12 0.28 0.00 1.22 0.36 0.42 < .01
Absences (1s
t
Quarter) -0.52 0.11 -0.39 0.00 -0.51 0.19 -0.37 0.01
Initial Interest -0.19 0.15 -0.14 0.22 -0.26 0.15 -0.19 0.08
Initial Utility Value 0.19 0.14 0.14 0.18 0.15 0.14 0.11 0.28
Mastery-approach Goals 0.31 0.17 0.23 0.08 0.28 0.17 0.21 0.10
Performance-approach
Goals
0.09 0.13 0.06 0.50 0.14 0.13 0.11 0.27
Teacher Dummy Codes
Teacher 1 -0.55 0.35 -0.15 0.12 -0.60 0.48 -0.44 0.21
Teacher 2 -0.72 0.37 -0.20 0.05 -0.18 0.50 -0.13 0.73
Teacher 3 -0.35 0.40 -0.09 0.39 -0.54 0.55 -0.40 0.33
Teacher 4 -1.06 0.46 -0.21 0.02 0.11 0.69 0.08 0.87
Teacher 5 -0.72 0.34 -0.21 0.04 -0.33 0.45 -0.25 0.46
Interactions with Group
Group * Perf. Exp. -0.51 0.23 -0.19 0.03
Group * Absences 0.11 0.24 0.08 0.66
Group * Sex 0.44 0.47 0.16 0.35
Group * White -0.80 0.50 -0.28 0.11
Teacher 1 * Group 0.09 0.70 0.06 0.90
Teacher 2 * Group -1.06 0.69 -0.78 0.13
Teacher 3 * Group 0.57 0.77 0.42 0.46
Teacher 4 * Group -1.77 0.90 -1.30 0.05
Teacher 5 * Group -0.89 0.65 -0.66 0.17
R2 .47** .58**
R2 - change .10*
Note: b = unstandardized regression coefficient. SE = standard error. β = standardized regression
coefficient.
** p < .01. * p < .05.

Hulleman Relevance in Science 23

Table S6
Multiple regression results for future plans
b SE β p b SE β p
Intercept 3.02 0.21 0.00 2.95 0.18 0.00
Group 0.22 0.30 0.10 0.46 0.11 0.25 0.05 0.67
Performance Expectations 0.13 0.11 0.12 0.24 0.06 0.09 0.05 0.54
Sex 0.21 0.18 0.10 0.23 0.06 0.15 0.03 0.66
Asian 0.47 0.33 0.13 0.15 0.24 0.27 0.07 0.38
Black 0.02 0.25 0.01 0.94 -0.01 0.21 0.00 0.96
Hispanic 0.02 0.35 0.00 0.96 0.02 0.29 0.01 0.93
Initial Interest 0.46 0.09 0.42 0.00 0.19 0.08 0.18 0.01
Initial Utility Value 0.09 0.08 0.08 0.27 0.02 0.07 0.02 0.79
Mastery-approach Goals 0.08 0.09 0.07 0.38 0.04 0.07 0.04 0.56
Performance-approach
Goals
0.13 0.07 0.12 0.07 0.10 0.06 0.09 0.10
Teacher Dummy Codes
Teacher 1 0.02 0.37 0.00 0.96 0.01 0.31 0.00 0.98
Teacher 2 0.25 0.32 0.07 0.43 0.09 0.26 0.02 0.74
Teacher 3 0.05 0.30 0.01 0.88 0.07 0.25 0.02 0.78
Teacher 4 -0.14 0.28 -0.05 0.63 0.24 0.23 0.09 0.30
Teacher 5 -0.68 0.34 -0.20 0.04 -0.34 0.28 -0.10 0.23
Teacher 6 -0.06 0.27 -0.02 0.82 0.12 0.23 0.04 0.60
Interactions with Group
Group * Perf. Exp. -0.23 0.13 -0.16 0.08 -0.10 0.11 -0.07 0.36
Group * Sex -0.26 0.25 -0.11 0.30 0.05 0.21 0.02 0.79
Group * Black -0.15 0.35 -0.04 0.67 0.10 0.29 0.02 0.74
Group * Asian -0.25 0.44 -0.05 0.57 -0.21 0.36 -0.04 0.56
Group * Hispanic -0.33 0.51 -0.06 0.52 -0.32 0.43 -0.05 0.46
Teacher 1 * Group -0.01 0.54 0.00 0.98 0.25 0.45 0.05 0.58
Teacher 2 * Group 0.00 0.46 0.00 0.99 0.27 0.38 0.05 0.47
Teacher 3 * Group 0.53 0.47 0.09 0.26 0.25 0.39 0.04 0.53
Teacher 4 * Group -0.72 0.38 -0.21 0.06 -0.71 0.31 -0.21 0.02
Teacher 5 * Group 0.54 0.44 0.13 0.22 0.38 0.37 0.09 0.31
Teacher 6 * Group -0.17 0.38 -0.05 0.65 -0.17 0.31 -0.05 0.60
Final Science Interest 0.63 0.06 0.58 0.00
R2 .38** .58**
R2 - change .20**
Note: b = unstandardized regression coefficient. SE = standard error. β = standardized regression
coefficient.
** p < .01. * p < .05.

Hulleman Relevance in Science 24

Table S7
Frequency of connections to student’s lives by experimental condition
Control Relevance
Quality of Connections
(i.e., Dosage) N % N %
0 114 90 35 26
1 10 9 68 51
2 2 1 31 23
Total 126 100 134 100
Mean 0.10 0.91
SD 0.31 0.67

Hulleman Relevance in Science 25

Table S8
Multiple regression results for proportion of days attended during the semester
Step 1 Step 2
b SE β p b SE β p
Intercept 0.95 0.02 0.00 0.97 0.02 0.00
Group 0.00 0.02 -0.02 0.85 -0.05 0.03 -0.24 0.18
Performance Expectations -0.01 0.01 -0.06 0.63 -0.02 0.02 -0.21 0.20
Sex -0.01 0.02 -0.07 0.44 -0.04 0.02 -0.20 0.10
Asian 0.05 0.05 0.09 0.31 0.09 0.06 0.17 0.13
Black -0.04 0.02 -0.18 0.06 -0.06 0.03 -0.26 0.05
Hispanic -0.05 0.05 -0.10 0.29 -0.07 0.07 -0.14 0.30
Initial Interest -0.01 0.01 -0.08 0.57 -0.01 0.01 -0.06 0.68
Initial Utility Value -0.01 0.01 -0.10 0.38 -0.01 0.01 -0.10 0.39
Mastery-approach Goals 0.01 0.01 0.12 0.40 0.01 0.01 0.11 0.43
Performance-approach
Goals
-0.01 0.01 -0.08 0.41 -0.01 0.01 -0.07 0.48
Absences (1s
t
Quarter) -0.04 0.01 -0.41 0.00 -0.03 0.01 -0.34 0.02
Teacher Dummy Codes
Teacher 1 0.00 0.03 0.01 0.94 -0.01 0.03 -0.05 0.74
Teacher 2 -0.08 0.03 -0.35 0.01 -0.10 0.03 -0.45 0.00
Interactions with Group
Group * Perf. Exp. 0.02 0.02 0.15 0.30
Group * Sex 0.05 0.04 0.21 0.21
Group * Black 0.04 0.04 0.11 0.39
Group * Asian -0.09 0.11 -0.09 0.41
Group * Hispanic 0.03 0.09 0.04 0.75
Group * Absences -0.01 0.02 -0.08 0.59
Teacher 1 * Group -0.01 0.04 -0.03 0.86
Teacher 2 * Group 0.13 0.05 0.37 0.01
R2 .34** .37**
R2 - change .04
Note: b = unstandardized regression coefficient. SE = standard error. β = standardized regression
coefficient. N = 108.
** p < .01. * p < .05.

Hulleman Relevance in Science 26

Appendix: Intervention Instructions

Control Group Unit Review Activity
Now that we have reviewed the main topics and concepts from this unit, it is time to reflect on one specific
topic or concept.
Part A: Pick one of the topics or concepts that we have covered in this unit.
Part B: Summarize main parts of this topic/concept.
You can either: 1) write about it in at least 5 sentences, 2) draw a concept map with a description, or 3)
draw a sketch with a description. If you do a concept map or a sketch, be sure to describe it well enough
so that the reader can understand it.
For example, if you were studying nutrition, you could choose a topic such as how food is digested. A
written summary would include a description of the digestive system, and how foods are broken down in
the mouth, stomach, and intestines. This process is called digestion. Food is broken down into
carbohydrates, proteins, and fats.
You could also draw a concept map of the digestive system. An example is provided below. Remember
that you would also need to add a brief written description with a concept map or diagram.
Digestive
S
y
stem Stomach Digestion
Intestines
Mouth
Fats
Proteins
Carbohydrates
Healthy
Remember: Do both Part A (pick a topic) and Part B (summarize the main parts).
The unit we are studying is: ______________________________________________
Part A: The topic/concept I pick is: _________________________________________
Part B: My summary and review (use the back side if needed):

Hulleman Relevance in Science 27

Relevance Group
Unit Review Activity
Now that we have reviewed the main topics and concepts from this unit, it is time to reflect on one specific
topic or concept.
Part A: Pick one of the topics or concepts that we have covered in this unit and briefly summarize the
main parts.
Part B: Apply this topic/concept to your life, or to the life of someone you know. How might the
information be useful to you, or a friend/relative, in daily life? How does learning about this topic apply to
your future plans?
You can either: 1) write about it in at least 5 sentences, 2) draw a concept map with a description, or 3)
draw a sketch with a description. If you do a concept map or a sketch, be sure to describe it well enough
so that the reader can understand it.
For example, if you were studying nutrition, you could choose a topic such as how food is digested.
Briefly summarize the digestive process—how foods are broken down in the mouth, stomach, and
intestines to make energy. Then you could write about how this applies to your own life. For example,
eating healthy foods helps your body produce energy to play your favorite sport or study for exams.
You could also draw a concept map of how your knowledge of digestion applies to your life. An example
is provided below. Remember that you would also need to add a brief written description with a concept
map or diagram.
Digestive
S
y
stem Stomach Energy
Intestines
Mouth
Sports
Studying
Healthy
Remember: Do both Part A (pick a topic and summarize) and Part B (apply it to life).
The unit we are studying is: _________________________________________________
Part A: The topic/concept I pick is: ____________________________________________
My brief summary:
Part B: My application to life (use the back side if needed):

Hulleman Relevance in Science 28


Supplemental References
1. J. M. Harackiewicz, A. M. Durik, K. E. Barron, E. A. Linnenbrink, J. M. Tauer, J Educ
Psy 100, 105-122 (2008).
2. C. S. Hulleman et al., J Educ Psy 100, 398-416 (2008).
3. S. W. Raudenbush, A. S. Bryk, Hierarchical linear models: Applications and data
analysis methods (Sage, Thousand Oaks, CA, ed. 3, 2002).
4. L. S. Aiken, S. G. West, Multiple regression: Testing and interpreting interactions (Sage,
Newbury Park, CA, 1991).
5. H. Bloom, Eval Rev 8, 225-246 (1984).
