Content uploaded by Andrew J Martin
Author content
All content in this area was uploaded by Andrew J Martin on Jan 12, 2017
Content may be subject to copyright.
PB Goal Setting and Achievement
0
Martin, A.J., & Elliot, A.J. (2016). The role of personal best (PB) goal setting in students’ academic
achievement gains. Learning and Individual Differences, 45, 222-227. DOI:
10.1016/j.lindif.2015.12.014.
This article may not exactly replicate the authoritative document published in the journal. It is not
the copy of record. The exact copy of record can be accessed via the DOI:
10.1016/j.lindif.2015.12.014.
PB Goal Setting and Achievement
1
SHORT REPORT
The Role of Personal Best (PB) Goal Setting in Students’ Academic Achievement Gains
Andrew J. Martin
School of Education, University of New South Wales, Australia
Andrew J. Elliot
Department of Clinical and Social Sciences in Psychology, University of Rochester
Requests for further information about this investigation can be made to Professor Andrew J.
Martin, School of Education, University of New South Wales, NSW 2052, AUSTRALIA. E-Mail:
andrew.martin@unsw.edu.au. Phone: +61 2 9385 1952. Fax: +61 2 9385 1946.
Thanks are extended to the Australian Research Council for funding and Educational Assessment
Australia (including Michelle O’Dowd, Dr Sofia Kesidou, Dr Rassoul Sadeghi, Nick Connolly,
Glynis Brown, Ardi Pradana, Sarah Loxton) for assistance with data collection.
PB Goal Setting and Achievement
2
SHORT REPORT
The Role of Personal Best (PB) Goal Setting in Students’ Academic Achievement Gains
Abstract
An experimental study was conducted to assess the role of personal best (PB) goal setting in gains
(or declines) in mathematics achievement. A total of 89 elementary and secondary school students
participated in a pre/post treatment/control group experimental design to test whether setting a
specific PB target score for an upcoming achievement test leads to achievement growth on that test.
The treatment group (PB goal setting) demonstrated greater achievement growth than the control
group between pre- and post-testing, including after controlling for mastery, performance-approach,
performance-avoidance, and test strategy goals. This study provides support for the proposition that
PB goal setting is associated with achievement growth in students’ academic lives.
Keywords: goal setting; personal best (PB) goals; academic growth; achievement; mathematics
PB Goal Setting and Achievement
3
SHORT REPORT
The Role of Personal Best (PB) Goal Setting in Students’ Academic Achievement Gains
In a climate of benchmarks, comparisons, accountability, and league tables, it is important to
ensure that students are not excluded from access to academic success or denied a sense of
academic progress (Anderman, Anderman, Yough, & Gimbert, 2010). Greater attention to
individuals’ academic growth may provide a foundation for giving a wide range of students a better
sense of their academic progress. The present research investigates growth by way of personal best
(PB) goal setting, and its role in academic achievement gains. PB goals are defined as specific,
challenging, competitively self-referenced targets to which students strive to match or exceed a
previous best. Examples of such targets include increased learning or better performance on current
schoolwork than in previous efforts (Martin, 2006, 2011; Martin & Liem, 2010; Yu & Martin,
2014). PB goals reside alongside other growth approaches to student development, such as value-
added models and the modeling of academic trajectories (Anderman et al., 2010; Harris, 2011).
Prior work into PB goals has been survey-based and correlational. The present study investigates a
PB goal setting intervention (having students set a PB goal) using an experimental design.
Prior Research on PB (and Other Growth) Goals
PB (and Other Growth) Goals: Correlational Work
A number of survey-based studies have demonstrated a connection between PB goals and
academic outcomes. In a cross-sectional study of high school students, Martin (2006) showed that
PB goals positively predicted students’ educational attainment aspirations, class participation,
enjoyment of school, and perseverance. In cross-lagged longitudinal work with high school
students, Martin and Liem (2010) found that PB goals predicted later literacy achievement,
numeracy achievement, effort on tests, perseverance, school enjoyment, class participation,
homework completion, educational attainment aspirations, and engagement. In a study focusing on
academically at-risk (attention-deficit/hyperactivity disorder; ADHD) students, Martin (2012) found
that the positive effects of PB goals generalized to students with ADHD. Following this, Yu and
PB Goal Setting and Achievement
4
Martin (2014) and Martin and Elliot (2014) examined PB goals alongside “classic” mastery and
performance goals among Chinese and Australian (respectively) middle and secondary school
students, finding a positive role for both PB and mastery goals. A study using a longitudinal cross-
lagged panel design found that high school students’ PB goals played a role in the development of
their implicit beliefs about intelligence, with PB goals positively predicting subsequent incremental
beliefs about intelligence and negatively predicting subsequent entity beliefs (Martin, 2014).
Other research has also investigated growth goals. Elliot, Murayama, and Pekrun (2012)
explored self-based goals. According to these researchers, “self-based goals use one’s own
intrapersonal trajectory as the evaluative referent” (p. 322). Interestingly, the statistically significant
findings for these goals were sparse; self-approach goals were significantly positively related to
approach temperament and feeling energized in class, but were not significantly related to several
other variables, including intrinsic motivation and achievement. More recent work by Elliot,
Murayama, Kobeisy, and Lichtenfeld (2014) explored past- and potential-based (growth-oriented)
goals, finding separability between these goals and a sound psychometric basis upon which to
explore their relationships with academic outcomes. Of note, in both studies, the goals that were
studied by Elliot and colleagues focused on exams (hence, a possible reason for equivocal findings),
but each of these goals -- self-based, past-based, and potential-based -- is equally applicable to other
activities and outcomes.
PB (and Other Growth) Goals: Experimental or Intervention Work
Very little work has been conducted on PB goal setting interventions. One recent study of PB
goals found that students in a PB goal setting treatment group for a self-paced science education
program reported significantly higher science aspirations at the end of the program, compared with
a no-goal control condition (Martin, Durksen, Williams, Kiss, & Ginns, 2014). There has also been
very little work investigating achievement gains following growth goal setting. In the earliest work
to our knowledge, Alschuler (1969) found that typing students setting personally challenging goals
aimed at faster typing speed through the course of their learning demonstrated a greater increase in
PB Goal Setting and Achievement
5
speed than a control group. Early work by Slavin (1980) was also promising. He had students set
individual targets that exceeded their prior level of achievement and provided rewards based on
improvement. Slavin found that over time, students in the treatment (growth target) group
outperformed students in a control group. However, in a follow-up study, Beady, Slavin, and
Fennesy (1981) failed to find such an effect, and it is our understanding that no subsequent work
has been conducted to better understand the discrepancy between these two studies. Thus, across
experimental growth goal setting designs there is a tendency to see educational gains, but this
finding should be considered tentative at present.
PB Goals: Conceptual and Applied Terrain
Achievement goal theory is one perspective relevant to the study of PB goals. At a fundamental
level, achievement goal theory is grounded in a distinction between mastery-approach goals focused
on understanding, developing skill, or improvement, and performance-approach goals focused on
outperforming others or demonstrating comparative competence (Elliot, 2005). Two other “classic”
goals include a mastery-avoidance goal (aiming to avoid misunderstanding and/or the loss of
knowledge or competence) and a performance-avoidance goal (aiming to avoid the demonstration
of incompetence relative to others and/or avoid poor performance in competitive or comparative
tasks). PB goals are distinct from performance goals in that the former are set in relation to self
(Martin, 2006), whereas the latter are set in relation to others (Elliot, 2005). PB goals may be
differentiated from mastery-approach goals in that mastery goals, as operationalized in the present
study (and elsewhere; e.g., see also Elliot & McGregor, 2001), are task-based (i.e., master a
mathematics task, learn a mathematics skill) and self-based (i.e., do better than one did on a
previous mathematics test; Vansteenkiste, Lens, Elliot, Soenens, & Mouratidis, 2014), whereas PB
goals are self-based alone. Notably, the recently proposed 3 x 2 achievement goal framework now
includes self-based (growth) goals along the lines of PB goals (Elliot, Murayama, & Pekrun, 2011).
We would also suggest that PB goals are highly dissimilar to mastery-avoidance and
performance-avoidance goals in that the latter two goals are avoidance oriented and focused on the
PB Goal Setting and Achievement
6
concept of incompetence (or a loss of competence). Of the two avoidance goals, performance-
avoidance tends to be emphasized in empirical research and may also be more salient and
recognizable in the classroom (Martin, 2013); it is therefore the performance-avoidance goal that is
included in our research. Taken together, in order to understand the unique effect of PB goal setting,
we include performance (approach and avoidance) and mastery-approach goals in analyses, so as to
partial out variance attributable to these goals and gain a better understanding of PB goals,
independent of “classic” achievement goals.
Goal setting frameworks (e.g., Locke & Latham, 2002) also provide useful insight into the
mechanisms by which PB goals may positively impact educational outcomes. Specifically, PB (and
other growth) goals may make it clear to students what they need to strive for to outperform a
previous best; PB goals may help students direct attention and effort towards the goal-relevant tasks
that are important to attain educational outcomes; through self-competition, PB goals may energize
students; and, PB goals may create a discrepancy between current and desired attainment, a gap that
students are motivated to close (Martin, 2011). Further, according to Senko, Hulleman, and
Harackiewicz (2011), goals that comprise challenging standards create an internal pressure to
perform, arouse energy and effort, and lead to success. On a related note, a meta-analysis by
Hulleman and colleagues (2010) found that challenge-seeking goals are more likely to predict
achievement than mastery or learning-oriented goals. While challenging goals can apply to both PB
and performance goals, we maintain that it is the personally-referenced challenge that is likely to be
more aligned to the intrinsically-motivated striving central to achievement growth (Martin, 2011).
What More Needs to be Known?
There are three gaps in the research base. First, no published research, to our knowledge, has
involved students setting a PB goal leading up to an achievement test and explored achievement
gains beyond possible gains made by students who set no such goal. Second, little intervention
research has investigated goal setting in a way that controls for individuals’ other achievement
goals. There is thus a need to explore the effects of a PB goal setting intervention controlling for the
PB Goal Setting and Achievement
7
presence of other goals (e.g., mastery and performance goals) that students may pursue. Third, and
finally, there is a need to differentiate any potential PB goal setting effect from students’ PB goal
orientation (i.e., students’ general or characteristic tendency to pursue PB goals). A significant PB
goal setting effect would suggest this as a successful intervention, irrespective of a students’ general
orientation to pursue PB – or other – goals.
Method
Sample
Participants were 89 elementary and secondary school students (25% non-government school,
75% government school; 88% co-educational, 8% single-sex boys, 4% single-sex girls) taking an
annual mathematics test administered by a scholastic assessment center based in a capital city on
the east coast of Australia. Students participated in 2012 and again in 2013. They were randomly
assigned, in a stratified manner (to optimize demographic equivalence), to a treatment or control
group. The treatment group comprised 41 students who set a PB goal (a score bettering their 2012
score) leading up to the 2013 test. The control group comprised 48 students who did not set a goal
leading up to the 2013 test. For the treatment group: 44% were female, 56% were male; 51% were
elementary school students, 49% were secondary school students; and the mean age was 12.32 (SD
= 1.85) years. For the control group: 44% were female, 56% were male; 48% were elementary
school students, 52% were secondary school students; and the mean age was 12.21 (SD = 1.86)
years. In terms of demographics, there were no significant differences between treatment and
control groups in gender representation, χ2(1) = .01, p = .99; school stage representation
(elementary, secondary school), χ2(1) = .10, p = .76; and age, t(87) = -.28, p = .78.
Procedure
The treatment group was informed that they were to set a PB target score for the 2013 test that
bettered their score on the 2012 test. Initial instructions were as follows: “Last year you scored xx /
40 in your mathematics test. Can we encourage you to set a Personal Best (PB) target for this year’s
mathematics test that is higher than last year’s test?” Then, treatment group participants were asked
PB Goal Setting and Achievement
8
to enter a PB target score as follows: “My Personal Best (PB) target for this year’s test is zz / 40.”
They were also instructed: “Make sure this PB target score is higher than last year’s score.”
Students’ 2013 target scores were then checked to ensure that this score exceeded their 2012 score.
Having entered their PB target score, they were informed: “It is important that you remember your
PB target score for the upcoming mathematics test. Now that you have set this PB target, can we
encourage you to remember this target as you do your preparation and the test?” To close the
treatment instruction session, treatment group students were instructed to print out a page that
contained the following details: “Last year I scored xx / 40 in my mathematics test. My Personal
Best (PB) target for this year’s mathematics test is: zz / 40.” Leading up to the test, the treatment
group was reminded (by e-mail and/or SMS text) of their PB target score four times (one week
prior, three days prior, one day prior, and on test day), as follows: “Just a reminder about your
Personal Best (PB) Target score. Last year, you scored xx / 40 on the Mathematics test. This year,
you set your PB Target at zz / 40 for the Mathematics test. To help remind you of this PB Target,
you will have printed out your Target score zz / 40 when you set it earlier this week. Try to
remember this PB Target leading up to the Mathematics test this week”. The control group received
no such reminders and no additional instructions; they were simply informed when the test would
be administered. Approximately four weeks following their invitation to participate, all treatment
and control group students took the 2013 test. The test was pencil-and-paper and conducted at
school. Schools and proctors administering the test were blind to the groups to which students were
assigned and to the specific nature and aims of the experiment.
Materials
The measures administered comprised mathematics achievement tests, goal intentions, and
basic demographics.
Mathematic achievement
The mathematics test comprised 40 items and was conducted once in August 2012 and again
in August 2013. The test suite (also comprising science, English, and computing options), is a well-
PB Goal Setting and Achievement
9
established and long-standing assessment tool used in over 6000 schools in Australia, and also
administered to students in over 20 countries (e.g., the US, Hong Kong, India, Malaysia, Singapore,
South Africa) each year. In Australia and New Zealand alone, for example, over 1 million tests are
taken each year. The test is taken by students who opt in (usually with the encouragement or
support from parents/carers) or who are part of an entire year group at a school that has decided to
take the test. In relation to the test under focus in this study, it is widely recognized as a valid and
objective measure of mathematics achievement that can complement or augment within-school
mathematics testing. For those opting in, it is likely a test that is taken seriously by the student
and/or his or her parents/carers, as it is generally seen as a means to gain further insight (beyond
curricular school assessment) into the student’s mathematics knowledge and skill-set. To the extent
that this is the case, we believe it is a test of sufficient high stakes for the students in both groups
(i.e., irrespective of goal setting or no goal setting) to work hard and care about the results.
All items are peer-reviewed by an external panel consisting of practising mathematics
teachers and mathematics experts, including university academics. Once approved by the panel,
questions are placed in tests in such a way as to balance item difficulty and the strands, topics, and
skills tested. The test items cover: algebra and patterns, chance and data, measures and units,
number and arithmetic, and space and geometry. For each grade level a different test was
administered in order to ensure appropriate test difficulty for different age groups and to enhance
test equivalence across grade levels. Thus, students completed parallel (not identical) tests in 2012
and 2013. To further ensure test equivalence, we standardized all treatment and control students’
test scores within students’ grade/year level (M = 0, SD = 1) separately in 2012 and again in 2013
(though for completeness, we also present raw score data in the Results). For elementary school
students, all 40 items were multiple choice and calculators were not permitted. For secondary
school students, 35 items were multiple choice, 5 items were free response, and calculators were
permitted (due to test difficulty). Means and other descriptive statistics for each group are presented
PB Goal Setting and Achievement
10
in Table 1. In the main, it is evident that the achievement measures were approximately normally
distributed as indicated by skewness and kurtosis values.
Goal intentions leading up to the test
A survey was also administered to participants. This was focused on their goal intentions
leading up to the 2013 test. The survey was administered once, after receiving the 2012 results and
prior to the PB goal setting manipulation for the 2013 test. Five goal intentions were assessed: PB
goals, mastery goals, performance-approach goals, performance-avoidance goals, and test strategy
goals. The items were preceded with the following stem: “In preparation for the mathematics test, I
will …”. All items were rated by students on a 1 (strongly disagree) to 7 (strongly agree) scale.
Although we have complete data on participants’ achievement for these analyses, surveys were
completed by 78% of participants. Importantly, however, there was no significant difference in non-
response between treatment and control groups, χ2(1) = .38, p = .53, nor in pre-test achievement,
t(87) = 1.33, p = .19, or post-test achievement, t(87) = .20, p = .84. We therefore conclude that
(survey) non-responders are not significantly different from responders.
PB goal intentions: To assess PB goal intentions, the four PB items in the Personal Best Scale
(Martin & Liem, 2010) were adapted (using the stem, “In preparation for the mathematics test, I
will …) as follows: “… try to get a better score than I did last year”; “… try to improve on the score
that I got last year”; “… try to get a better result than last year’s result”; and, “… try to beat the
score that I got last year”. Reliability for the PB goal intention scale was high for the treatment
group (Cronbach’s α = .93) and for the control group (Cronbach’s α = .99).
Mastery, performance-approach, and performance-avoidance goal intentions: To assess
mastery goal intentions, two mastery items most aligned with learning and mastery for a test
(reduced from the full item set due to time constraints) from the Achievement Goal Questionnaire
(Elliot & McGregor, 2001) were adapted (using the stem, “In preparation for the mathematics test, I
will …”) as follows: “…try to understand what I am studying as much as possible” and “…try to
master what I am studying as much as possible”. Reliability for mastery goals was good (treatment
PB Goal Setting and Achievement
11
group’s Cronbach’s α = .79; control group’s Cronbach’s α = .85). To assess performance-approach
and avoidance goal intentions, four items (reduced from the full item set due to time constraints)
from the Achievement Goal Questionnaire (Elliot & McGregor, 2001) were adapted (using the
stem, “In preparation for the mathematics test, I will …”) as follows: “…try to outperform most of
the other students” and “…try to do well compared to most of the other students” for performance-
approach; and, “…try not to fail the test” and “…try to avoid performing poorly on the test” for
performance-avoidance. Reliability was deemed acceptable for performance-approach goals
(treatment group’s Cronbach’s α = .96; control group’s Cronbach’s α = .64 [not unacceptably low
given an item-total correlation of .61 for this 2-item scale], and good for performance-avoidance
goals (treatment group’s Cronbach’s α = .86; control group’s Cronbach’s α = .73).
Test strategy goal intentions: Finally, a set of thirteen test strategy goal intention items were
asked of students. These were adapted from an inventory by Martin (2003) identifying self-
regulatory strategies that students can use in their preparation for upcoming tests and exams. Using
the same stem as other goal intentions (“In preparation for the mathematics test, I will …”), some
example items were as follows: “… study regularly”; “develop a study timetable and stick to it”;
“… look at past test papers or other practice material”; “…deal with distractions that can arise such
as friends, sports, hobbies, etc.”; and, “… try to get good sleep in the week leading up to the test”.
Reliability for this measure was good (treatment group’s Cronbach’s α = .92; control group’s
Cronbach’s α = .86).
Data Analysis
Given the research design and the research question, our analysis focused on determining the
extent and nature of gains or declines in achievement as a function of group (treatment = PB goal
setting, control = no goal setting) membership. Testing for the effect of PB goal setting on students’
achievement gains entailed four analyses of covariance (ANCOVA; see Winer, Brown, & Michels,
1991). ANCOVA has more power to test for control-treatment differences when using a
randomized design (Van Breukelen, 2006), as we do. The ANCOVA included 2012 achievement as
PB Goal Setting and Achievement
12
the covariate and 2013 achievement as the dependent measure, conducted in hierarchical fashion as
follows: main effect of group (treatment, control); main effect of group, gender, and stage
(elementary, secondary school); main effect of group, gender, stage, and their 2-way interactions;
and, main effect of group, gender, stage, their 2-way interactions, and their 3-way interaction.
Central assumptions for the conducting of ANCOVA were met
1
. To test for any differences in prior
goal intentions, t-test procedures were employed. Additionally, the five goal intentions were
included as covariates in a subsequent ANCOVA to test for the effects of the PB goal setting
manipulation beyond variance explained by goal intentions.
Results
Testing for PB Goal Setting
After establishing that there was no significant difference in pre-test achievement between the
treatment (PB goal setting) group and the control group, t(87) = -.52, p = .60, we conducted four
ANCOVAs, in hierarchical fashion. First, we entered group (treatment, control) as a main effect,
2012 achievement as the covariate, and 2013 achievement as the dependent variable. There was a
significant treatment effect such that the PB goal group demonstrated a significantly greater
achievement gain than the no-goal control group between 2012 and 2013, F(1,86) = 10.05, p < .01.
Second, we entered group (treatment, control), gender, and stage (elementary, secondary
school) as main effects, 2012 achievement as the covariate, and 2013 achievement as the dependent
variable. There was a significant treatment effect such that the PB goal group demonstrated a
significantly greater achievement gain than the no-goal control group, F(1,84) = 10.00, p < .01.
There were no significant findings for the main effects of gender and stage.
Third, we entered group, gender, and stage as main effects, their 2-way interactions (i.e.,
group x gender, group x stage, gender x stage), 2012 achievement as the covariate, and 2013
1
The dependent variable and covariate were continuous; the dependent variable was approximately normally
distributed for each category of the independent variable (see Table 1 SDs, skewness and kurtosis); and there was
homogeneity of variance for groups on the achievement pre-test covariate (F = .01, p = .92) and post-test dependent
variable (F = .28, p = .60). Particularly for ANCOVA, two additional assumptions were met: for each level of the
grouping variable, the covariate was similarly linearly related to the dependent variable (control, r = .65, p < .001;
treatment, r = .77, p < .001); and, there was homogeneity of regression slopes as indicated by a non-significant pre-test
x group interaction for the post-test score, F = .56, p = .81.
PB Goal Setting and Achievement
13
achievement as the dependent variable. With the inclusion of the 2-way interactions, there remained
a significant treatment effect such that the PB goal group demonstrated a significantly greater
achievement gain than the no-goal control group, F(1,81) = 11.47, p < .001. There were no
significant findings for the other main effects of gender and stage and also no significant effects for
the 2-way interactions.
Finally, we conducted a 3-way ANCOVA with 2012 achievement as the covariate and 2013
achievement as the dependent measure. As the full and final model, ANCOVA results are presented
in Table 2. The patterns of means are presented in Figures 1a and 1b. Again, there was a significant
treatment effect such that the PB goal group demonstrated a significantly greater achievement gain
than the no-goal control group between 2012 and 2013, F(1,80) = 11.22, η2p = .12, p < .001
2
. The
effect size for pre-test differences between groups was small (d = .11). The difference between
groups in post-test achievement was medium to large (d = .56; Cohen, 1992). In addition, this main
effect was not qualified by interactions with gender or school stage.
Testing for Potential Effects of Other Goals
As a further test of the PB goal setting condition, beyond the effects of other possible goals
held by the two groups, the 2 (group) x 2 (gender) x 2 (stage) ANCOVA (with 2012 achievement as
the covariate and 2013 achievement as the dependent measure) was once again conducted, but this
time including the five goal intentions (PBs, mastery, performance-approach, performance-
avoidance, test strategy) as covariates. This partials out the effect of goal intentions on 2013
achievement, enabling a closer assessment of the unique effects of the PB goal setting condition.
Again, there was a significant treatment effect such that beyond variance explained by goal
intentions (and prior achievement, gender, and stage), the PB goal setting group demonstrated
significantly greater achievement gain than the no-goal control group between 2012 and 2013,
F(1,59) = 16.53, p < .001.
2
For completeness, we conducted a 2 (control, treatment) x 2 (female, male) x 2 (elementary, secondary school) x 2
(2012, 2013 test) ANOVA with repeated measures on the last factor. This yielded a significant group (control,
treatment) x time (2012, 2013) interaction, F(1,81) = 7.04, MS = 1.90, p < .01 such that the treatment (PB goal) group
made significant gains between 2012 and 2013 relative to the control (no PB goal) group.
PB Goal Setting and Achievement
14
Discussion
All previous PB goal research has been correlational and predominantly self-reported and
focused on “intrapsychic” academic outcomes in the form of motivation and engagement. The
present study involved an experimental goal setting task relevant to achievement. This represents a
major extension of previous research, and the positive finding observed for PB goals further
supports the validity of the concept. Indeed, the fact that the PB goal setting condition produced a
main effect that was not qualified by interactions with gender or stage of school, and was not
diminished with inclusion of various goal intentions as covariates, suggests its generality across
males and females, different developmental groups, and different goal orientations (although
additional, high powered studies would be needed before a definitive conclusion could be drawn in
this regard). One recent study has shown that a PB goal setting intervention leads to gains in
students’ educational aspirations (Martin et al., 2014). The present study has extended this
experimental program by showing gains in achievement.
The goal setting manipulation was the only goal factor shown to distinguish the two groups in
a statistically significant manner. The fact that the groups did not statistically vary on goal
intentions relevant to PBs, mastery, performance-approach, performance-avoidance, and test
strategy goals – but did statistically vary in their PB goal setting – suggests that there is a PB goal
setting effect beyond (other) specific goal intentions held by the groups (although here again
additional studies would be needed before a definitive conclusion could be drawn). This was further
supported by the subsequent ANCOVA that demonstrated a PB goal setting effect, beyond variance
explained by the five goal intentions as covariates. Thus, although we cannot rule out both groups
holding specific goals leading up to the test, we can assert that the PB goal setting activity does
differentiate the two groups, providing a basis for attributing effects to the PB goal setting activity.
Beyond both groups’ PB (and other) goal intentions, the effect of actually setting a PB target
leading up to a test seems to be beneficial. This, we maintain, is an important contribution showing
that beyond goal intentions, goal setting plays an important role in students’ academic growth.
PB Goal Setting and Achievement
15
Guided by Yu and Martin (2014), there are numerous practical strategies for potentially
amplifying the effect of PB goal setting on students’ achievement growth. For example, educators
may encourage students to consider both process- and outcome-oriented PB goals that are
specifically directed at improving achievement. Process PB goals that may promote achievement
might include such aims as spending extra time revising for an upcoming test than a previous test,
asking a teacher for help in test preparation when normally the teacher would be avoided,
maintaining a more organized study environment than was the case for previous tests, showing
more or better workings in one’s mathematics calculations, spending more time checking grammar
and punctuation for a written assessment task, and attempting more questions in an upcoming test
than in the previous test. Outcome PB goals might include performing better on this week’s test
than last week’s test, getting more sums correct in one’s mathematics quiz, and performing better
on end-of-year exams compared to half-yearly exams (Martin, 2011; see also
www.lifelongachievement.com for PB goal worksheets for students and teachers).
Our review of the literature in the introduction section identified a number of conceptual
perspectives that have some alignment and relevance with PB goals, and that also represent
something of a conceptual backdrop and inspiration for growth goal research. The findings from
this study suggest opportunities for further research and thinking with regards to growth and related
theory. In terms of goal setting theory (Locke & Latham, 2002), one line of questioning that the
present data were not able to inform is the precise PB goal mechanism/s that account/s for academic
growth. For example, goal setting theory articulates the role of task specificity, task clarity,
difficulty, challenge, personal reference, and the like. Which of these dimensions explain relatively
greater variance in academic growth? Further work would be helpful here.
With respect to achievement goal theory, we focused predominantly on approach goals in our
work. The achievement goal literature clearly attests to the conceptual and predictive utility of the
approach-avoidance distinction, and it may be useful to apply this distinction to the PB concept as
well (for relevant research, see Elliot et al., 2011, 2014). In relation to implicit beliefs about
PB Goal Setting and Achievement
16
intelligence and growth mindsets (Dweck, 2006, 2012), what is the relative salience and causal
ordering of PB goals and implicit beliefs with respect to academic motivation, engagement, and
achievement? Preliminary longitudinal work by Martin (2014) suggests, perhaps counter-
intuitively, that PB goals may actually shape the development of students’ incremental beliefs about
intelligence.
Limitations and Future Directions
This study provides insight into the role of PB goal setting in predicting gains in achievement.
In the context of the positive findings for PB goals, it is important to note some limitations, so as to
contextualize these findings. The first limitation is that the study would be enhanced by a larger
sample and an additional treatment condition that taught students how to pursue the PB goals that
they set. Second, it would also be helpful to include a treatment condition that instructs students to
set other goals, such as performance-approach goals, to determine the predictive role of PB goals
relative to these other goals. Related to this, given our work focused only on performance-
avoidance goals, future research might include mastery-avoidance goals to identify the role of PB
goals relative to this additional aspect of the 2x2 achievement goal framework. In addition, with
respect to performance-avoidance goals, we make the point that our operationalization focused on
avoiding poor test performance and not avoiding poor performance relative to others, which has
been a suggested operationalization in prior reviews of goal constructs (e.g., Hulleman et al., 2010).
To better align with our performance-approach operationalization, this comparative “other” referent
might be included in future research. Third, although we included measures to differentiate PB goal
setting from specific goal intentions, there is a need to more directly confirm that the control group
did not set a PB (or similar) goal and that it is the growth nature of the goal setting task that yielded
the effect, not simply goal specificity. Fourth, assigning a PB goal to students might not be as
effective as students self-selecting such a goal. It is unclear how many students would have actually
chosen a PB goal, if given the option. Finally, there is also the question of the role of PB goals in
PB Goal Setting and Achievement
17
relation to other recently proposed growth-oriented goals, such as potential-based goals (Elliot et
al., 2014).
PB Goal Setting and Achievement
18
References
Alschuler, A.S. (1969). The effects of classroom structure on achievement motivation and
academic performance. Educational Technology, IX, 19-24.
Anderman, E., Anderman, L., Yough, M., & Gimbert, B. (2010). Value-added models of
assessment: Implications for motivation and accountability. Educational Psychologist, 45,
123-137. doi.org/10.1080/00461521003703045
Anderman, E.M., & Wolters, C.A. (2006). Goals, values, and affect: Influences on student
motivation. In P.A. Alexander & P. Winne. (Eds). Handbook of educational psychology (pp.
369-389). Mahwah, NJ: Erlbaum.
Beady, C., Slavin, R., & Fennessey, G. (1981) Alternative student evaluation structures and a
focused schedule of instruction in an inner-city junior high school. Journal of Educational
Psychology, 73, 518-523. doi.org/10.1037/0022-0663.73.4.518
Carver, C.S., & Scheier, M.F. (1996). Perspectives on personality. Needham Heights, MA: Allyn
& Bacon.
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159. doi:10.1037/0033-
2909.112.1.155.
Dweck, C.S. (2006). Mindset: The new psychology of success. New York: Random House.
Dweck, C.S. (2012). Mindsets and human nature. American Psychologist, 67, 614-622
Elliot, A.J. (2005). A conceptual history of the achievement goal construct. In A.J. Elliot & C.S.
Dweck (Eds). Handbook of competence and motivation. (pp. 52-72). New York: Guildford.
Elliot, A.J., & McGregor, H. A. (2001). A 2 × 2 achievement goal framework. Journal of
Personality and Social Psychology, 80, 501-519. doi.org/10.1037/0022-3514.80.3.501.
Elliot, A.J., Murayama, K., Kobeisy, A., & Lichtendfeld, S. (2014). Potential-based achievement
goals. British Journal of Educational Psychology. doi:10.1111/bjep.12051
Elliot, A.J., Murayama, K., & Pekrun, R. (2011). A 3 x 2 achievement goal model. Journal of
PB Goal Setting and Achievement
19
Educational Psychology, 103, 632-648. doi.org/10.1037/a0023952
Ford, M.E. (1992). Motivating humans: Goals, emotions, and personal agency. Newbury Park, CA:
Sage. doi.org/10.4135/9781483325361
Harris, D.N. (2011) Value-added measures in education. Cambridge: Harvard Educational Press.
Hattie (2009). Visible learning. New York: Routledge.
Hattie (2012). Visible learning for teachers. New York: Routledge.
Hulleman, C.S., Schrager, S.M., Bodmann, S.M., & Harackiewicz, J.M. (2010). A meta-analytic
review of achievement goal measures: Different labels for the same constructs or different
constructs with similar labels? Psychological Bulletin, 136, 422-449.
doi.org/10.1037/a0018947
La Guardia, J.G., & Ryan, R.M. (2002). What adolescents need: A self-determination theory
perspective on development within families, school, and society. In F. Pajares & T. Urdan
(Eds). Academic motivation of adolescents. Connecticut: Information Age Publishing.
Liem, G.A.D., & Martin, A.J. (2013). Direct instruction and academic achievement. In J. Hattie &
E. Anderman (Eds.). International Guide to Student Achievement. Oxford: Routledge.
Locke, E.A., & Latham, G.P. (2002). Building practically useful theory of goal setting and task
motivation. American Psychologist, 57, 705-717. doi.org/10.1037/0003-066X.57.9.705
Maehr, M.L., & Zusho, A. (2009). Achievement goal theory: The past, present, and future. In K.R.
Wentzel & A. Wigfield (Eds). Handbook of motivation at school. (pp. 77-104). New York:
Routledge.
Martin, A.J. (2003). How to motivate your child for school and beyond. Sydney: Bantam.
Martin, A.J. (2006). Personal bests (PBs): A proposed multidimensional model and empirical
analysis. British Journal of Educational Psychology, 76, 803-825.
doi.org/10.1348/000709905X55389
Martin, A.J. (2011). Personal best (PB) approaches to academic development: Implications for
PB Goal Setting and Achievement
20
motivation and assessment. Educational Practice and Theory, 33, 93-99.
doi.org/10.7459/ept/33.1.06
Martin, A.J. (2012). The role of Personal Best (PB) goals in the achievement and behavioral
engagement of students with ADHD and students without ADHD. Contemporary
Educational Psychology, 37, 91-105. doi.org/10.1016/j.cedpsych.2012.01.002
Martin, A.J. (2013a). Goal orientation. In J. Hattie & E. Anderman (Eds.). International Guide to
Student Achievement. Oxford: Routledge.
Martin, A.J. (2013b). Improving the achievement, motivation, and engagement of students with
ADHD: The role of personal best goals and other growth-based approaches. Australian
Journal of Guidance and Counselling, 23, 143-155. doi.org/10.1017/jgc.2013.4
Martin, A.J. (2014). Implicit theories about intelligence and growth (personal best) goals:
Exploring reciprocal relationships. British Journal of Educational Psychology.
doi:10.1111/bjep.12038.
Martin, A.J., Durksen, T., Williamson, D., Kiss, J., & Ginns, P. (2014). Personal best (PB) goal
setting and students’ motivation in science: A study of science valuing and aspirations.
Australian Educational and Developmental Psychologist, 31, 85-96.
Martin, A.J., & Elliot, A.J. (2014). The role of Personal Best (PB) and “classic” achievement goals
in students’ academic growth: A longitudinal investigation of gains and declines in motivation
and engagement. Submitted for publication.
Martin, A.J., & Liem, G.A. (2010). Academic personal bests (PBs), engagement, and achievement:
A cross-lagged panel analysis. Learning and Individual Differences, 20, 265-270.
doi.org/10.1016/j.lindif.2010.01.001
Senko, C., Hulleman, C.S., & Harackiewicz, J.M. (2011). Achievement goal theory at the
crossroads: Old controversies, current challenges, and new directions. Educational
Psychologist, 46, 26-47. doi.org/10.1080/00461520.2011.538646
PB Goal Setting and Achievement
21
Shalem, Y., Sapire, I., & Huntley, B. (2013). Mapping onto the mathematics curriculum – an
opportunity for teachers to learn. Pythagoras, 34, Art. #195, 1-10.
http://dx.doi.org/10.4102/pythagoras.v34i1.195
Slavin, R. (1980). Effects of individual learning expectations on student achievement. Journal of
Educational Psychology, 72, 520-524. doi.org/10.1037/0022-0663.72.4.520
Van Breukelen, G.J.P. (2006). ANCOVA versus change from baseline had more power in
randomized studies and more bias in nonrandomized studies. Journal of Clinical
Epidemiology, 59, 920-925.
Vansteenkiste, M., Lens, W., Elliot, A. J., Soenens, B., & Mouratidis, A. (2014). Moving the
achievement goal approach one step forward: Toward a systematic examination of the
autonomous and controlled reasons underlying achievement goals. Educational Psychologist,
49, 153-174.
Winer, B.J., Brown, D.R., & Michels, K.M. (1991). Statistical principles in experimental design.
New York, NY: McGraw-Hill.
Yu, K., & Martin, A.J. (2014). Personal best (PB) and ‘classic’ achievement goals in the Chinese
context: Their role in predicting academic motivation, engagement, and buoyancy.
Educational Psychology: An International Journal of Experimental Educational Psychology,
34, 635-658. doi.org/10.1080/01443410.2014.895297
PB Goal Setting and Achievement
22
Table 1. Achievement Raw Means, SDs, Skewness, Kurtosis by Group (Control, Treatment), Gender
and Stage (Elementary, Secondary School)
M
SD
Skewness
Kurtosis
2012
2013
2012
2013
2012
2013
2012
2013
CONTROL
Elementary
- Females
28.75
26.50
5.97
5.73
-1.49
-.14
2.41
-1.19
- Males
30.18
29.36
4.35
4.50
-.62
-.44
-.15
.41
Secondary
- Females
27.44
27.00
5.85
6.16
-.29
1.21
-.76
1.85
- Males
31.06
30.19
5.54
5.90
.28
.19
-1.16
-1.60
Total
29.60
28.48
5.47
5.67
-.48
.17
.38
-.71
TREATMENT
Elementary
- Females
29.91
32.27
4.70
4.19
-.62
.01
.83
.25
- Males
30.70
32.50
5.17
5.02
-.57
-.79
-.53
-.08
Secondary
- Females
29.71
30.71
5.31
5.99
-.49
-.45
.60
-1.31
- Males
29.54
29.70
5.62
7.06
-1.17
-1.33
1.41
1.26
Total
29.95
31.24
5.04
5.65
-.73
-1.11
.17
1.43
PB Goal Setting and Achievement
23
Table 2. Analysis of Covariance: Effect of PB Goal Setting on Achievement Gains
Note. PB = personal best
df
Mean Square
F
Partial eta
square
Sig
2012 Math Score (pre-test; covariate)
1
31.24
71.67
.47
<.001
Group (Control/Treatment)
1
4.89
11.22
.12
<.001
Gender (Female/Male)
1
.19
.43
.01
.52
Stage (Elementary/Secondary)
1
.02
.04
.01
.85
Group x Gender
1
.94
2.16
.03
.15
Group x Stage
1
.97
2.21
.03
.14
Gender x Stage
1
.05
.11
.01
.74
Group x Gender x Stage
1
.02
.04
.01
.84
PB Goal Setting and Achievement
24
Figure 1a. Standardized (by grade level; M = 0, SD = 1 for each grade level) achievement scores
for treatment and control groups, 2012-2013. (PB = personal best)
Figure 1b. Raw achievement scores (out of 40) for treatment and control groups, 2012-2013. (PB =
personal best)
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
2012 Achievement 2013 Achievement
Control (No PB Goal)
Treatment (PB Goal)
25
26
27
28
29
30
31
32
33
34
35
2012 Achievement 2013 Achievement
Control (No PB Goal)
Treatment (PB Goal)
Achieve
Score
Achieve
Score
