Computers & Education 152 (2020) 103869

Contents lists available at ScienceDirect

Computers & Education

journal homepage: http://www.elsevier.com/locate/compedu

Online vs traditional homework: A systematic review on the

benefits to students’ performance
~es *, 1, Diogo Ferreira 1, Jennifer Cunha , Pedro Rosa
Paula Magalha �rio
Escola de Psicologia, Universidade do Minho, Portugal

A R T I C L E I N F O A B S T R A C T

Keywords:
Applications in subject areas
Distance education and online learning
Teaching/learning strategies
In recent years, a strong investment has been made to equip schools with technology for
educational purposes, leading to some transformations in teachers’ practices. In this context,
apart from traditional format, there are online platforms available for teachers to assign and
monitor homework completed by students. Literature has described the benefits of online
homework for teachers when compared with traditional homework. Still, evidence of the benefits
for students when compared with traditional homework is scattered. Thus, the current study
aimed to conduct a systematic review on online homework and traditional homework, with the
goal of learning the methodological features of the studies comparing both homework formats
and understanding which of the two benefits most students’ performance. Following a review
protocol, 31 studies were included and categories such as participants, duration of intervention,
and course domain were described. Results showed that most studies used a convenience sample
in a between-subject design, included college students, used non-standardized instruments, and
presented a wide range of number of participants, duration of intervention, course domain, and
approaches to implement the online homework. Additionally, half of the studies showed no
differences in terms of student performance as a function of the format of delivery of homework.
In nine studies students’ performance was higher in the online compared with the traditional
homework format, and only one study showed the opposite result. Lastly, six studies reported
mixed results. Based on results, educational implications are discussed, as well as suggestions for
future research.

1. Introduction

Literature considers homework as a facilitator tool for students’ engagement and academic performance (e.g., Cooper, Robinson, &
Patall, 2006; Hagger, Sultan, Hardcastle, & Chatzisarantis, 2015; author, 20xx). Nevertheless, homework is not effective per se; on the
contrary, homework efficacy depends on its characteristics (Dettmers, Trautwein, Lüdtke, Kunter, & Baumert, 2010; Rosa �rio, Nún~ ez,
Vallejo, Nunes, Cunha, Fuentes, & Valle, 2018). Throughout the last decades, several characteristics of homework (e.g., frequency,
purpose, amount; see Cooper, 2001) have been extensively analyzed (see Keith, 1982; Foyle, Lyman, Tompkins, Perne, & Foyle, 1990;
Ros�
ario, Cunha, Nunes, Nunes, Moreira, & Nún ~ ez, 2019; Trautwein, 2007); still, the format of homework delivery has been less
investigated. Considering the increased use of technology in today’s classrooms, namely the assignment of online homework (e.g.,

* Corresponding author. School of Psychology, University of Minho, Campus de Gualtar, 4710-057, Braga, Portugal.
E-mail address: pmagalhaes@psi.uminho.pt (P. Magalh~ aes).
1
These authors contributed equally to this work and should be cited as co-first authors.

https://doi.org/10.1016/j.compedu.2020.103869
Received 28 March 2019; Received in revised form 28 January 2020; Accepted 8 March 2020
Available online 12 March 2020
0360-1315/© 2020 Elsevier Ltd. All rights reserved.
P. Magalh~
aes et al. Computers & Education 152 (2020) 103869

Dufresne, Mestre, Hart, & Rath, 2002), it seems relevant to study homework format (i.e. paper-and-pencil and online). For this reason,
the current study aims to investigate via a systematic review the methodological features of the studies comparing both homework
formats and understand which of two homework formats benefits most students’ performance.

1.1. Traditional homework: research findings

Homework is an educational tool present in classrooms from distinct cultures and is assigned to, and completed by, students
irrespective of grade level, subject or course (see Chang, Wall, Tare, Golonka, & Vatz, 2014; Cooper, Lindsay, Nye, & Greathouse, 1998;
Trautwein & Lüdtke, 2007). This tool can be defined as tasks, assigned to students by their teachers, that are intended to be carried out
during non-school hours (Cooper, 2001). Homework provides teachers the opportunity to extend the time of learning outside school
hours and students the occasion to review, practice, and consolidate what they have learned in class (i.e. instructional purposes), as
well as to monitor students’ learning and difficulties (Cooper, 2001; Ros� ario, Cunha, Nunes, Moreira, Nú~ nez, & Xu, 2019; Epstein &
Van Voorhis, 2001). Typically, teachers, especially from elementary and middle school, assign homework in the traditional format (e.
g., choosing exercises from the students’ textbooks; Kaur, 2011; Rosa �rio et al., 2019), which has been designated as traditional
homework (Dufresne et al., 2002), and will be used from now onwards.
A substantial amount of research conducted in the past few decades has focused on the contribution of traditional homework to
students’ engagement and academic achievement (Buijs & Admiraal, 2013; Nún ~ ez, Sua
�rez, Cerezo, Gonza�lez-Pienda, Ros� ario, Mour~ao,
& Valle, 2013; Ros� ario et al., 2018; see also the meta-analyses conducted by; Cooper et al., 2006; Fan, Xu, Cai, He, & Fan, 2017). Still,
taken together, data on homework indicates both positive and negative impacts on students’ learning. Literature shows that
completing homework assignments has positive effects for students of different school levels. Benefits include higher self-efficacy
beliefs, self-regulatory skills (e.g., planning), engagement in school (e.g., pay attention, class participation), and conscientiousness
(e.g., being careful, hardworking, self-controlled) (Buijs & Admiraal, 2013; Galloway, Conner, & Pope, 2013; Go €llner et al., 2017;
Ros� ario, Mour~ao, Baldaque, Nunes, Nún ~ ez, Gonza�lez-Pienda, & Valle, 2009). However, existing literature also alerts to the fact that
homework may also have negative effects on students, especially when its characteristics are not balanced (e.g., amount, degree of
difficulty, instructional purpose). Negative effects include procrastination behaviors, negative emotions, physical fatigue, and pressure
to complete homework and outperform, which may lead to copied responses from other students (Dettmers et al., 2011; Fulano, Cunha,
Nún ~ ez, Pereira, & Rosa�rio, 2018; Galloway et al., 2013; Trautwein, Niggli, Schnyder, & Lüdtke, 2009). Despite this, homework is
perceived as a valuable teaching tool that fosters academic achievement and success. Students who complete their homework on a
regular basis tend to achieve higher grades than their counterparts who do not complete their assignments (Fan, Xu, CaiHe, & Fan,
2017; Fern� andez-Alonso, Alvarez-Díaz,
� Sua�rez-Alvarez,
� & Mun ~ iz, 2017).
Besides teachers’ involvement in planning the design of homework tasks (Epstein & Van Voorhis, 2001; Rosa �rio et al., 2019), they
also play an important role in providing feedback on the homework handed in by students (Cooper, 2001; Nún ~ ez, Sua
�rez, Rosa �rio,
Vallejo, Cerezo, & Valle, 2014). The latter is important due to the positive impact homework feedback plays on student’s achievement,
especially when homework is graded or commented on (Ros� ario, Nún ~ ez, Vallejo, Cunha, Nunes, Su� arez, & Moreira, 2015). Teachers’
feedback is considered the key to maximizing the positive impact of homework on students’ learning and achievement (Walberg &
Paik, 2000). However, as teachers’ involvement in administrative tasks mounts, grading traditional homework may be a burden and
often considered a non-feasible task (Cunha et al., 2018; Ros� ario et al., 2019). This heavy workload is likely to prevent delivering
on-task feedback adjusted to students’ learning needs, which may thwart students from truly benefitting from it (Cunha et al., 2019;
Demirci, 2010a; Shute, 2008). The use of technology in class, particularly of homework systems that provide students with instant
feedback, seems to offer an alternative for teachers to overcome these barriers (Stickles, 2017).

1.2. Technology and online homework: research findings

Computers and portable devices are entrenched in our everyday lives. Nowadays, technology provides motivated students the
ability to learn and access information on-demand, without geographic limitations, whenever they are curious about a subject (Chan
et al., 2006).
Technological advances are visible in today’s classroom. In fact, there is fast growth in the use of computers, web-based tools, and
even digital games for teaching and learning (e.g., Lim, Zhao, Tondeur, Chai, & Tsai, 2013; Proctor & Marks, 2013). These tools help
teachers deal with the increasing number of students per class and with the burden of their administrative workload (Elmehdi,
Ibrahim, & Haba, 2013). For example, regarding homework, the use of an online or web-based format has clear benefits for teachers (e.
g., reduction in the time needed to copy, distribute, and collect homework; less time to check the assignments; provision of on-time and
personalized feedback) when compared with the traditional format (Ismail, Mokhtar, Nasir, Rashid, & Ariffin, 2014). The use of an
online tool has the cost of setting up the system, teaching the students how to use it and how to download information, and answering
student questions regarding software issues. However, the system also allows teachers to substantially reduce the time spent grading
the homework assignments of all the students, monitor their learning status, and adjust teaching strategies accordingly to respond to
students’ needs (Fynewever, 2008; Lunsford & Pendergrass, 2016).
The use of web platforms to assign homework also has many benefits for students. Students can receive immediate and individ­
ualized feedback, which provides them the opportunity to correct their errors, submit new versions of the responses to the assignments
in a timely manner, and may encourage students to practice the exercises to achieve mastery (e.g., Halcrow & Dunnigan, 2012;
Rodriguez & Smith, 2016; Lunsford & Pendergrass, 2016). In fact, students appreciate the possibility to learn instantly their homework
grade and be provided with hints to help them find the correct answer, which is only possible using an online homework system (Davis

2
P. Magalh~
aes et al. Computers & Education 152 (2020) 103869

& McDonald, 2016). This translates in positive relationship between online homework completion and final grade (e.g., Eichler &
Peeples, 2013; Parker & Loudon, 2013). Lastly, these platforms allow the randomization of exercises which is likely to reduce student’s
dishonest behaviors related to copying homework from their peers (Bonham, Beichner, & Deardorff, 2001).
Despite the aforementioned benefits, online homework also has its drawbacks. For example, trial-and-error submissions due to the
possibility of multiple submissions may entice students to guess the answer (Kortemeyer, 2015), which emphasizes the final answer
rather than the process of building answers (Bonham et al., 2001; Mendicino, Razzaq, & Heffernan, 2009).

1.3. Motivation for the study

Across OECD countries, on average, 48% of students reported using a computer to do homework, 38% to using e-mail to
communicate about schoolwork with their peers, and 33% to sharing school-related materials via computer (OECD, 2014). The current
generation of students use technology in their daily life, and some complain that completing traditional homework is boring. Thus,
asking students to use a computer for homework assignments might foster their school and homework engagement (Dillard-Eggers,
Wooten, Childs, & Coker, 2008). Literature suggests that positive attitudes towards homework are associated with higher achievement
(Chang et al., 2014). Similarly, students’ motivation towards homework is affected by the way they perceive their ability to learn
through homework (Smith & McCoy, 2011). Besides the argument that using technology to complete homework tasks may foster
engagement in the activity, the possibility of providing instant feedback to students should also be considered as an important aspect in
which technology may be beneficial to students. Literature on traditional homework shows that, when teachers grade homework and
mark incorrect answers accompanied by an explanation on how they can improve and correct those mistakes, students are likely to
improve the quality of their work (Elawar & Corno, 1985; Williams, 2010). Instant feedback as provided by the online platforms may,
therefore, contribute to students’ learning.
In a scenario of extensive use of technology, it becomes important to systematize research findings regarding the benefits of
delivering homework in one format over the other. Existing literature reports unequivocally the benefits of online homework for
teachers when compared with traditional homework (e.g., reduction in the time needed to copy, distribute, and collect homework, less
time to check the assignments). Still, evidence of online homework enhanced benefits for students when compared with traditional
homework is, at best, scattered. Likewise, it is important to understand the specific methods of delivery and execution of each
homework format in the different grade levels as research findings may not generalize to all developmental stages.
This systematic review aims to help practitioners, researchers, and educational systems developers consolidate knowledge in the
literature and also the research and design gaps that could be addressed in future developments. Thus, the following research questions
guided the current study:
Q1: Which are the methodological features (participants, course domain, study design, characteristics of the interventions, online
platform, and homework assignments) of the studies comparing online and traditional homework?
Q2: Which format of homework (online or traditional) shows more benefits for student’s performance?

2. Materials and methods

The present systematic review focused on the comparison of online homework and traditional homework in terms of its benefits for
students’ performance. Cochrane Collaboration (Higgins & Wells, 2011) guidelines were followed in the development of the manu­
script, including the search, the selection, and the extraction of the material, as well as the guidelines of The Preferred Reporting Items
for Systematic Reviews and Meta-Analysis (PRISMA) (Moher, Liberati, Tetzlaff, & Altman, 2009).

2.1. Search strategy and selection criteria

A systematic literature search was performed with a preplanned review protocol specifying inclusion and exclusion criteria for the
studies and subsequent data analysis. Only original research, peer-reviewed, and English-language studies were considered, excluding
review articles and short conference abstracts. Inclusion criteria were as follows: (1) the study had to compare online and traditional
homework regarding their effect on students’ performance, and (2) the study needed to include a quantitative measure allowing for
that comparison (e.g., final grades). That is, papers were excluded when only one format of homework was assessed (online or
traditional) or when the only form of measurement was students or instructors’ perceptions about each format of homework. Three
databases were searched, namely Web of Science, Education Resources Information Centre (ERIC), and Google Scholar. The keywords
used for the searches were “online homework” and “web-based homework”. These keywords were chosen because “online homework”
is the term authors use to define the format of homework and “web-based homework” was chosen because after the first database
search the articles also included this keyword as descriptor. The searches were conducted between November 6th, 2017 and April 16th,
2018. For the Web of Science database, authors used the keywords as Topic and refined the search to show only articles. As for ERIC,

Table 1
Search results and final articles selected per database.
Database Web of Science Google Scholar ERIC

Results 67 317 122

Articles Selected 6 12 13

3
P. Magalh~
aes et al. Computers & Education 152 (2020) 103869

authors used the keywords described above. Finally, for Google Scholar, authors used the tool “advanced research” and wrote the
keywords on “with all the words” and “with the exact expression”, authors also ticked the option “on the tittle of the article”. Table 1
shows the search results and papers included for each database.

2.2. Selection process and data extraction

Authors followed PRISMA (Moher et al., 2009) guidelines for the data abstraction. The PRISMA Statement provides an evidence-
based 27-item checklist (e.g., on objectives, methodology, limitations) for systematic reviews and meta-analyses.
The team included DF, PM, and JC. DF screened the titles and abstracts. Afterwards, two researchers (DF and PM) independently
conducted the eligibility phase. In case of uncertainty or disagreement between DF and PM, a third researcher (JC) refereed the
discussion (i.e. DF and PM reviewed all papers and, when in doubt regarding a paper, JC screened the full text to later discuss it with
the team). Consensus was reached through discussion. DF and PM followed the inclusion and exclusion criteria described in section
2.1. for the inclusion through the abstract. Afterwards, data was extracted as follows: participants, course, study design, characteristics
of the intervention, characteristics of the online platforms, and study findings. Finally, DF extracted all relevant information for the
tables and both PM and JC independently read all the included papers and reviewed the extracted data.

Fig. 1. PRISMA Flow Diagram: Flow of information through the search and selection process during this systematic review.

4
P. Magalh~
aes et al. Computers & Education 152 (2020) 103869

2.3. Data analysis

For the purpose of this review and to facilitate reading and understanding of the results, authors grouped papers in four course
domains. The groups were as follows: Mathematics, including papers with a focus on algebra and statistics; Business, comprising papers
on accounting, business statistics, and economics; Sciences, including papers on physics, chemistry, and environmental sciences; and
the last group comprising papers focused on Engineering. Acknowledging that some articles used the academic year, instead of se­
mesters, to describe the duration of the study, for the purpose of this investigation we considered two semesters equal to an academic
year and considered one term as one semester. Concerning the online homework tool used in each study, data was reported as
“textbook developed”, when the platform was developed by a textbook publisher; “developed by the researchers”, when the platform
was purposely developed for the study; “university developed”, when the platform was developed by a university; or “company
developed”, when the platform was developed by a company not necessarily a textbook publisher. Finally, regarding the results and
conclusions of the articles, for the purpose of the current research authors opted to report data as: Positive, when online homework had
better results than traditional homework; Neutral, when results were similar in both formats of homework; and Negative, when online
homework had poorer results than the traditional homework.

3. Results

The database searches resulted in 507 records, of which 112 were duplicates with 395 articles remaining. After the title selection,
261 studies remained, and abstracts were screened for eligibility. At this stage, 97 studies were excluded by abstract screening.
Therefore, 164 full-texts were retrieved for assessment against the inclusion and exclusion criteria and 133 did not comply with the
criteria. Finally, 31 individual studies were included in the final review. Fig. 1 outlines these results by displaying the flow of in­
formation throughout the different phases of the systematic review.

3.1. Methodological features of the studies

The dates of the publications of the 31 studies included in this systematic review ranged from 2002 to date; however, the majority
was published after 2009 (see Fig. 2). The primary aim of the reviewed studies was to assess whether the performance of students (e.g.,
homework grades, final grade in the course) improved by using an online homework platform when compared to counterparts using
traditional homework. In the following sections the main characteristics of the selected studies were described, including participants,
course, study design, characteristics of the intervention, and characteristics of the online platforms and homework delivered. When
possible, information was synthetized into tables as follows: information regarding the academic year, course domain, and study
design per article was organized in Table 2; and information regarding the characteristics of the online platforms and homework per
article was arranged in Table 3.

3.1.1. Participants
The majority of the articles included did not provide complete information about their participants, making it difficult to draw a
clear demographic profile. Still, except for four studies (Chow, 2014; Davis & McDonald, 2016; Dodson, 2014, Dufresne et al., 2002),
the remaining 27 provided information about the size of their sample. When taken together, the size of the samples of the 27 studies
ranged from 52 (Burch & Kuo, 2010) to 442 (Jonsdottir, Bjornsdottir, & Stefansson, 2017) participants, with a mean number of 181
participants (SD ¼ 112).
Thirteen studies specified the gender of the participants (Bonham, Deardorff, & Beichner, 2003; Brewer & Becker, 2010; Demirci,
2007; Elmehdi et al., 2013; Fatemi, Marquis, & Wasan, 2014; Fratto, Sava, & Krivacek, 2016; Gaffney, Ryan, & Wurst, 2010; Hauk,
Powers, & Segalla, 2014; Jonsdottir et al., 2017; Malik, Martinez, Romero, Schubel, & Janowicz, 2014; Mendicino et al., 2009;
Williams, 2010; Woolley, 2015). There was a mean of 102 (SD ¼ 91) female participants, ranging from 24 (Fratto et al., 2016) to 303
(Hauk et al., 2014). For the males, the mean number of participants was 85 (SD ¼ 58), ranging from 29 (Elmehdi et al., 2013) to 186

Fig. 2. Number of publications per year since the first paper, of the included publications, addressing the online homework topic was published.

5
P. Magalh~
aes et al. Computers & Education 152 (2020) 103869

Table 2
Course domain and study design per article.
Academic Course Study Design
Year
Convenience Randomized Between- Within- Follow- Pre and post
sample sample subject subject up measure

Arora et al. (2013) College Mathematics X X X X

Babaali and College Mathematics X X
Gonzalez (2015)
Bonham et al. (2003) College Sciences X X
Brewer and Becker College Mathematics X X X
(2010)
Burch and Kuo College Mathematics X X X
(2010)
Chow (2014) College Mathematics X X
Davis and McDonald College Engineering X X
(2016)
Demirci (2007) College Sciences X X
Demirci (2010a) College Sciences X X X
Demirci (2010b) College Sciences X X X
Dodson (2014) High School Sciences X X
Dufresne et al., 2002 College Sciences X X
Elmehdi et al. (2013) College Sciences X X
Fatemi et al. (2014) College Business X X
Fratto et al. (2016) College Business X X X
Fynewever (2008) College Sciences X X X
Gaffney et al. (2010) College Business X X
Gok (2011) College Sciences X X X
Hauk et al. (2014) College Mathematics X X X
Johnston (2004) College Business X X
Jonsdottir et al. College Mathematics X X
(2017)
Lee et al. (2010) College Business X X X
Lenz (2010) College Mathematics X X
Malik et al. (2014) College Sciences X X
Mathai and Olsen College Mathematics X X
(2013)
Mendicino et al. Fifth Grade Mathematics X X X
(2009)
Palocsay and Stevens College Business X X
(2008)
Rodriguez and Smith College Business X X
(2016)
Stickles (2017) College Mathematics X X
Williams (2010) College Mathematics X X X
Woolley (2015) College Business X X

(Gaffney et al., 2010). Only two studies included information about the age of the participants: in one study the mean age was 29.9
years (Williams, 2010); the other study presented data in intervals, in which the majority of the participants had between 18 and 24
years old and only two participants were between 25 and 34 years old (Fratto et al., 2016).
Twenty-nine studies were conducted with college students (see Table 2). Regarding these studies, 21 mentioned that participants
were undergraduate students but did not specify the course year in which students were enrolled. Still, four studies reported that
students were from the first year (Babaali & Gonzalez, 2015; Elmehdi et al., 2013; Hauk et al., 2014; Lenz, 2010), one study was
conducted with second-year students (Davis & McDonald, 2016), and another study included participants of the third and fourth year
(Rodriguez & Smith, 2016). The study by Malik et al. (2014) included students from the third and fourth year and non-degree students.
Finally, the study by Williams (2010) included master and doctoral students.
Lastly, two studies were conducted with non-college students. Particularly, in Dodson’s (2014) work, participants were students
from the third and fourth year of high school, whereas in the study by Mendicino and colleagues’ (2009) participants were fifth
graders.

3.1.2. Course domain

The studies were conducted in the following course domains: Mathematics, with 12 studies; Sciences, with 10 studies; Business,
with 8 studies; and Engineering, with 1 study. This information is displayed on Table 2.

3.1.3. Study design

Regarding the study design, except for one article (Jonsdottir et al., 2017) all articles used a convenience sample. By convenience
sample is meant the following: i) the researchers used data available from previous years (i.e. datasets) or ii) the sample was comprised

6
 P. Magalh~
Table 3

aes et al.
General characteristics of the Online Platforms per article and description of the different formats of questions and answers supported by the Online Platforms as described in each study. When authors
describe the specific formats of the answers adopted, it will be signaled. Otherwise, information refers solely to the capabilities of the platforms.
Developer Homework Feedback Answers format

Origin Type #errors MC NA SE TW SA T/F M DG DD WE WF UF ND

Arora et al. (2013) TB BE P ND X X X X

Babaali and Gonzalez (2015) TB DE P DT X X
Bonham et al. (2003) TB TE P; E ND X Xb
Brewer and Becker (2010) TB DE P NL X Xa Xa Xa
Burch and Kuo (2010) TB DE P TT X X
Chow (2014) TB BE P ND X X
Davis and McDonald (2016) ND ND P SD – X
Demirci (2007) ND BE P; E ND X X
Demirci (2010a) DR DE P; E ND X X
Demirci (2010b) DR DE P; E ND X X
Dodson (2014) DR ND Q ND – Xb Xb
Dufresne et al., 2002 TB DE P NL X X
Elmehdi et al. (2013) TB BE P NL X X X X X X X
Fatemi et al. (2014) TB DE P; E TT X X
Fratto et al. (2016) ND DE Q ND X X
b
Fynewever, 2008 TB BE P NL X X
Gaffney et al. (2010) TB DE P; E NL X X
Gok (2011) UD DE P; E ND X Xa Xa Xa Xa Xa Xa
7

Hauk et al. (2014) UD DE E ND X X

Johnston (2004) CD DE Q ND X Xa
Jonsdottir et al. (2017) OS DE E ND X Xb
Lee et al. (2010) CD DE P; Q ND X X
Lenz (2010) TB DE TT X X X X X X
Malik et al. (2014) TB ND P ND X Xa
Mathai and Olsen (2013) TB ND P TT X Xa Xa Xa
Mendicino et al. (2009) UD ND P ND X Xb Xb
Palocsay and Stevens (2008) TB DE P; Q ND X X X X X X X X
Rodriguez and Smith (2016) TB DE P ND X Xb
Stickles (2017) TB ND P ND X X
Williams (2010) TB ND P; Q ND X Xa Xa
Woolley (2015) TB ND P FT – X

Computers & Education 152 (2020) 103869

Notes.
Developer: TB - Text Book Publisher; DR – Developed by the Researchers; UD – University Developed; CD – Company Developed; OS – Open-source; ND - Not Defined.
Homework Origin: TE - Teacher Exercises; DE – Database Exercises; BE – Both types of Exercises.
Homework Type: P – Problems; E Exercises; Q – Questions.
Homework # errors limited: TT - Three Tries; FT - Five Tries; SD - Score Decreases with Tries; DT – Defined by the Teacher; NL – No limit; ND - Not Defined.
Answers format: MC - Multiple-choice; NA - Numerical answer; SE - Symbolic (algebraic/mathematical) expression/answer; TW - Typing a word; SA - Short answer/essay; T/F - True or false; M –
Matching; DG - Draw graphs; DD - Draw free-body diagrams or Structures; WE - Write equations; WF - Write formulas; UF - Upload a file; ND - Not described.
a
Authors describe the specific format of homework answers adopted.
b
Authors present screenshots or a description of sample homework exercises to illustrate the type of answer format.
P. Magalh~
aes et al. Computers & Education 152 (2020) 103869

Table 4
Study results per article. Articles are organized by course domain.
Course Domain Positive Negative Neutral Students’ Authors Conclusions(a)
Perceptions towards the online
RK EF HC OP PP RK EF HC OP PP RK EF HC OP PP
towards system
the Online HW

College Arora et al. (2013) X X X þ Favours online

Mathematics Babaali and Gonzalez X n/a Favours online
(2015)
Brewer and Becker X n/a Favours online
(2010)
Burch and Kuo X þ Favours online
(2010)
Chow (2014) X n/a Favours online
Hauk et al. (2014) X n/a Favours online
Jonsdottir et al. X þ Favours online
(2017)
Lenz (2010) X X þ Favours online
Mathai and Olsen X n/a Favours combining
(2013) traditional and online
formats
Stickles (2017) X X þ Favours online
Williams (2010) X X X X n/a Takes no position
College Sciences Bonham et al. (2003) X X X n/a Favours online
Demirci (2007) X X þ Favours online
Demirci (2010a) X X n/a Favours online
Demirci (2010b) X X þ Favours online
Dufresne et al., 2002 X X n/a Favours online
Elmehdi et al. (2013) X ¼ Favours online
Fynewever (2008) X X þ Favours online
Gok (2011) X X X n/a Takes no position
Malik et al. (2014) X þ Favours online
College Business Fatemi et al. (2014) X n/a Favours combining
traditional and online
homework
Fratto et al. (2016) X þ Favours online
Gaffney et al. (2010) X X ¼ Favours online
Johnston (2004) X þ Favours online
Lee et al. (2010) X n/a Takes no position
Palocsay and Stevens X n/a Takes no position
(2008)
Rodriguez and Smith X þ Favours online
(2016)
Woolley (2015) X þ Takes no position
Others Mendicino et al. X n/a Favours online
(2009)
Davis and McDonald X n/a Favours combining
(2016) traditional and online
homework
Dodson (2014) X X X n/a Favours online

Notes: Positive - Online homework better results than the traditional homework; Negative - Online homework worse
results
than the traditional
homework; Neutral – Similar results for both formats of homework; RK – Retention of Knowledge; EF – Exam and
Final Grades; HC - Homework
assignments completed; OP - Overall performance; PP – Pre and Post Test.
þ
Students’ opinions and perceptions favor the online homework format.
¼
Students’ opinions and perceptions are similar for both homework formats.
n/a
The study does not evaluate students’ perceptions and opinions about the homework format.
a
Authors of each study conclusion to the question regarding which format of homework (online or traditional)
shows more benefits for
student’s performance or the homework format that they recommend.

by the students enrolled in the subject which the researchers were teaching. Regarding the Jonsdottir et al. (2017) study, authors split
the class randomly into two groups; each completing the assignment using a particular format, traditional or online. After completing
the assignment, the groups took a test and swapped the homework format assigned. This cycle was repeated four times throughout the
study. Moreover, except for four studies (Elmehdi et al., 2013; Jonsdottir et al., 2017; Mendicino et al., 2009; Woolley, 2015), the
remaining employ a between-subject design, that is, students were assigned to one group only, traditional or online. Finally, 12 studies

8
P. Magalh~
aes et al. Computers & Education 152 (2020) 103869

used a pre-post-test measure, with only one study including a follow-up (Arora, Rho, & Masson, 2013). This information is displayed on
Table 2.

3.1.4. Intervention’s characteristics

The duration of the intervention across studies had a mean of 2.4 semesters (SD ¼ 1.5) and a median of two semesters (Arora et al.,
2013; Bonham et al., 2003; Burch & Kuo, 2010; Demirci, 2010a; Fatemi et al., 2014; Fratto et al., 2016; Malik et al., 2014; Mathai &
Olsen, 2013; Stickles, 2017). The shortest intervention lasted for four days (this last information was not used in the calculation of the
mean) (Mendicino et al., 2009) and the longest intervention lasted for five semesters (Lenz, 2010).

3.1.5. Online platforms’ characteristics

Table 3 summarizes the general characteristics of the online platforms used. Of the 31 studies included, 19 used platforms
developed by a textbook publisher. Nine used an online platform that was either purposively developed for the study, or developed by a
university, or developed by a company, or an open-source platform. Finally, three studies did not specify the origin of their online
platform. Twenty-eight studies explicitly mentioned that the platforms used included some kind of feedback that varied in terms of
complexity – from a correct vs. incorrect response (e.g., Fratto et al., 2016; Gaffney et al., 2010; Hauk et al., 2014) to a tutor-based
step-by-step feedback (e.g., Arora et al., 2013; Burch & Kuo, 2010; Fynewever, 2008; Mendicino et al., 2009). Platforms also
differed in terms of the possibility of the teachers including their own exercises besides the ones available on the platform (e.g., Arora
et al., 2013; Palocsay & Stevens, 2008). Two studies mention that students were able to print the exercises from the platform and solve
them off the computer (Gok, 2011; Hauk et al., 2014). Moreover, some studies mention that there were limited number of tries students
could attempt before submitting their homework (e.g., Fatemi et al., 2014; Lenz, 2010). Other studies did not limit the number of
homework attempts (e.g., Fynewever, 2008; Gaffney et al., 2010), whereas others included different types of limitations to the number
of attempts. For instance, Davis and McDonald (2016) reduced the scores students could get on a single homework as a function of the
number of attempts used to find the correct answer, whereas in the study by Babaali and Gonzalez (2015) the number of tries was
defined by the teacher and could vary for each question. Still, many studies did not describe whether the platform allowed to control
this aspect.

3.1.6. Homework’s characteristics

Regarding the characteristics of the homework prescribed, not much information was provided. The majority of the studies (25) do
not refer the purpose for assigning homework (Bonham et al., 2003; Brewer and Becker, 2010; Burch & Kuo, 2010; Chow, 2014; Davis
& McDonald, 2016; Demirci, 2007; Demirci, 2010a,b; Dodson, 2014; Dufresne et al., 2002; Elmehdi et al., 2013; Fatemi et al., 2014;
Fratto et al., 2016; Fynewever, 2008; Gaffney et al., 2010; Gok, 2011; Jonsdottir et al., 2017; Rodriguez & Smith, 2016; Lee, Courtney,
& Balassi, 2010; Lenz, 2010; Malik et al., 2014; Mendicino et al., 2009; Palocsay & Stevens, 2008; Stickles, 2017; Williams, 2010).
Nevertheless, among these studies, few systematize the general purposes for ascribing homework in that course domain, particularly: i)
practice problem-solving skills (Bonham et al., 2003; Davis & McDonald, 2016; Demirci, 2007, 2010a; Palocsay & Stevens, 2008); ii)
practice skills and concepts (Brewer and Becker, 2010; Dodson, 2014; Williams, 2010); iii) mastery of the material/deepening the
understanding (Demirci, 2010b; Dodson, 2014; Fratto et al., 2016); iv) introduction to new topics/learn concepts (Dodson, 2014;
Fratto et al., 2016); v) motivate and guide learning (Demirci, 2010b; Fynewever, 2008); and vi) increased time studying and practicing
(Gok, 2011). Conversely, six studies clearly identify the purpose to the homework assigned in their studies: practice problem-solving
(Arora et al., 2013); mastery-oriented exercises (Babaali & Gonzalez, 2015); skill building and practice (Hauk et al., 2014; Mathai &
Olsen, 2013); encourage students to read the textbook and keep on-track with the course materials (Johnston, 2004); and practice
concepts and increase test-taking abilities (Woolley, 2015).
Regarding the general characteristics of the homework assigned, Table 3 also synthesizes some information provided by the studies.
The main type of homework assigned, 25 studies classified it as problems. Regarding the format of the answer of the homework, 15
studies did not specify the format of the answer (Babaali & Gonzalez, 2015; Burch & Kuo, 2010; Chow, 2014; Davis & McDonald, 2016;
Demirci, 2007; Demirci, 2010a, b; Dufresne et al., 2002; Fatemi et al., 2014; Fratto et al., 2016; Gaffney et al., 2010; Hauk et al., 2014;
Lee et al., 2010; Stickles, 2017; Woolley, 2015).
Four studies (Arora et al., 2013; Elmehdi et al., 2013; Lenz, 2010; Palocsay & Stevens, 2008) describe what the online platforms
support in terms of format of the questions and answers, but authors of these studies did not specify which of the answers format was
adopted. Six studies (Brewer & Becker, 2010; Gok, 2011; Johnston, 2004; Malik et al., 2014; Mathai & Olsen, 2013; Williams, 2010)
explicitly enunciate the format of the answer to their homework assignments. Lastly, six studies (Bonham et al., 2003; Dodson, 2014;
Fynewever, 2008; Jonsdottir et al., 2017; Mendicino et al., 2009; Rodriguez & Smith, 2016) include examples of homework assign­
ments (e.g., screenshots, description). Still, it is not clear for the readers whether the examples provided represent the sole format of
question and answer adopted by the author or whether there were other formats of answers included in the studies.
In the majority of the studies, both traditional and online homework exercises were the same or very similar (Arora et al., 2013;
Babaali & Gonzalez, 2015; Bonham et al., 2003; Brewer & Becker, 2010; Demirci, 2007, 2010a, 2010b; Dodson, 2014; Dufresne et al.,
2002; Elmehdi et al., 2013; Fatemi et al., 2014; Fratto et al., 2016; Fynewever, 2008; Gok, 2011; Hauk et al., 2014; Jonsdottir et al.,
2017; Lenz, 2010; Malik et al., 2014; Mathai & Olsen, 2013; Mendicino et al., 2009; Williams, 2010). Particularly, the majority of the
studies refer that in both conditions, homework consisted of end-of-chapter assignments. Three papers described that students were
asked in the online homework condition to do their homework assignments individually, while students in the traditional condition
were asked to complete it in group (Demirci, 2007; 2010a; 2010b). Lastly, Babaali and Gonzalez (2015) mentioned that students in the
online group completed more homework assignments than the traditional group.

9
P. Magalh~
aes et al. Computers & Education 152 (2020) 103869

3.2. Main results

Next, main results of the studies included in this review will be analyzed. First, we will look at the overall picture, analyzing what
each study reported regarding efficacy of the online versus traditional homework format in relation to the benefits for students’
performance, and for each variable examined. Second, we will look at these results, particularly at college level, as a function of the
course domain of each study.

3.2.1. Benefits for students’ performance

When comparing online homework and traditional homework, the authors of the individual studies assessed and compared stu­
dents’ performance by using several types of measures. Specifically, retention of knowledge, exam and final grades, homework as­
signments completed, overall performance, and pre-post-test measure were used as measures of student performance in the different
studies.
To facilitate the interpretation of the results, the study findings of each individual study were organized in three categories:
Positive, Neutral, and Negative. Table 4 displays the study findings per article and per variable included in each study. Nine studies
were categorized only as Positive, meaning that the online homework group had better results in all the measures selected than the
traditional homework group (Arora et al., 2013; Babaali & Gonzalez, 2015; Burch & Kuo, 2010; Dufresne et al., 2002; Elmehdi et al.,
2013; Fratto et al., 2016; Jonsdottir et al., 2017; Malik et al., 2014; Mendicino et al., 2009). The majority of these studies used exams,
final grades, or homework assignments completed as measures. Moreover, 15 studies were categorized as Neutral, meaning that both
the online and traditional groups had similar results in the assessment (Bonham et al., 2003; Brewer & Becker, 2010; Chow, 2014;
Davis & McDonald, 2016; Demirci, 2010a; Fatemi et al., 2014; Fynewever, 2008; Hauk et al., 2014; Johnston, 2004; Lee et al., 2010;
Mathai & Olsen, 2013; Palocsay & Stevens, 2008; Rodriguez & Smith, 2016; Stickles, 2017; Woolley, 2015). These studies had a
greater diversity in terms of measures selected to assess efficacy, but the majority used the exam or final grade measure or a pre-post-
test. Finally, one study was categorized as Negative (Demirci, 2007), meaning that the online homework group had poorer results in
the measures selected (homework completion and overall performance) than the traditional homework group.
Six studies reported mixed results, either positive, negative or neutral, as a function of the variable being studied. Particularly,
Demirci (2010b) reported negative results for the variable homework completion and neutral results for the variable pre- and post-test.
Williams (2010) reported the exact same trend for the variables homework completion (negative) and pre- and post-test (neutral), with
the remaining variables (exam or final grade and overall performance) of this study being neutral. In contrast, for the variable homework
completion both Gaffney et al. (2010) and Gok (2011) reported positive results. These two studies also reported neutral results for the
variable exam or final grade, with Gok (2011) studying a third variable (pre- and post-test) also with a neutral result. Lenz (2010) re­
ported positive results for homework completion but negative results for exam or final grade. Finally, Dodson (2014) reported positive
results for exam or final grade, negative results for the variable overall performance, and neutral results for homework completion.

3.2.2. College course domain

To deepen the understanding of the main results, studies were grouped by course domain, and results were interpreted from there.
The goal is to learn if any conclusion can be drawn as a function of the course domain of the studies. Therefore, three domains were
analyzed (Mathematics, Sciences, and Business) and three individual studies were left out of this analysis: the two studies using a non-
college sample (Dodson, 2014; Mendicino et al., 2009) and the study on engineering (Davis & McDonald, 2016). Table 4 summarizes
this information.

3.2.2.1. Mathematics. The course domain Mathematics counts with 11 studies, of which five showed neutral results (Brewer & Becker,
2010; Chow, 2014; Hauk et al., 2014; Mathai & Olsen, 2013; Stickles, 2017), four had positive results (Arora et al., 2013; Babaali &
Gonzalez, 2015; Burch & Kuo, 2010; Jonsdottir et al., 2017), and two reported mixed results (Lenz, 2010; Williams, 2010). Five studies
did not describe the format of the answer of the online homework (Babaali & Gonzalez, 2015; Burch & Kuo, 2010; Chow, 2014; Hauk
et al., 2014; Stickles, 2017). Additionally, seven studies assessed only one variable, of which five used the variable exam and final
grades, three reported positive results (Babaali & Gonzalez, 2015; Burch & Kuo, 2010; Jonsdottir et al., 2017) and two neutral results
(Brewer & Becker, 2010; Mathai & Olsen, 2013). Two studies evaluated two variables (Lenz, 2010; Stickles, 2017), the first found
mixed results and the second reported neutral results; one study evaluated three variables and found positive results (Arora et al.,
2013); and one study evaluated four variables and found mixed results (Williams, 2010).
Regarding feedback, all studies explicitly described the type of feedback provided. Only one study (Hauk et al., 2014) reported
using corrective feedback, specifically correctness of the answer, with neutral results. All the remaining studies described more
complex formats of feedback, focusing on instructional aspects. Examples include: customized coaching and individualized feedback,
feedback on common errors, hints while the student completes the assignment, or description of how to solve the problem correctly.
Moreover, five studies assessed students’ perceptions towards the online homework format (Arora et al., 2013; Burch & Kuo, 2010;
Jonsdottir et al., 2017; Lenz, 2010; Stickles, 2017). All were positive towards the online format and preferred it over the traditional
homework format. Lastly, except for two studies (Mathai & Olsen, 2013; Williams, 2010), authors of the other nine studies concluded
favorably in relation of using online homework as the preferred format of homework delivery, even when their studies did not report
differences in terms of students’ performance.

3.2.2.2. Sciences. The course domain Sciences counts with nine studies, of which three showed neutral results (Bonham et al., 2003;

10
P. Magalh~
aes et al. Computers & Education 152 (2020) 103869

Demirci, 2010a; Fynewever, 2008), three showed positive results (Dufresne et al., 2002; Elmehdi et al., 2013; Malik et al., 2014), one
reported negative results (Demirci, 2007), and two reported mixed results (Demirci, 2010b; Gok, 2011). Four studies did not describe
the format of the answer of the online homework (Demirci, 2007, 2010a, 2010b; Dufresne et al., 2002). Additionally, except for two
studies (Elmehdi et al., 2013; Malik et al., 2014), all assessed more than one variable. Particularly, five included two measures
(Demirci, 2007, 2010a, 2010b; (Dufresne et al., 2002); Fynewever, 2008) and two included three measures (Bonham et al., 2003; Gok,
2011).
Regarding feedback, all studies explicitly described the type of feedback provided. Five studies (Bonham et al., 2003; Demirci,
2007, 2010a, 2010b; Malik et al., 2014) reported to use corrective feedback, specifically correctness of the answer and/or grading. Of
these studies, two reported neutral results (Bonham et al., 2003; Demirci, 2010a), one showed negative results (Demirci, 2007), one
showed mixed results (Demirci, 2010b), and the last one reported positive results (Malik et al., 2014). The remaining four studies
described more elaborated types of feedback, including suggestions of solution, step-by-step hints, or tailored feedback. Of these, two
reported positive results (Dufresne et al., 2002; Elmehdi et al., 2013), one showed neutral results (Fynewever, 2008), and one mixed
results (Gok, 2011). Moreover, five studies assessed students’ perceptions towards the online homework format (Demirci, 2007,
2010b; Elmehdi et al., 2013; Fynewever, 2008; Malik et al., 2014). While one study found that students did not prefer one homework
format over the other (Elmehdi et al., 2013), the other four studies reported that students preferred the online format over the
traditional homework format. Lastly, authors from these nine studies, except for the study by Gok (2011), concluded favorably in
relation of using online homework as the preferred format of homework delivery, even when their studies did not report differences in
terms of students’ performance.

3.2.2.3. Business. The course domain Business counts with eight studies, of which six reported neutral results (Fatemi et al., 2014;
Johnston, 2004; Lee et al., 2010; Palocsay & Stevens, 2008; Rodriguez & Smith, 2016; Woolley, 2015), one reported positive results
(Fratto et al., 2016), and one mixed results (Gaffney et al., 2010). Five studies did not describe the format of the answer of the online
homework (Fatemi et al., 2014; Fratto et al., 2016; Gaffney et al., 2010; Lee et al., 2010; Woolley, 2015). Additionally, these eight
studies, except for one (Gaffney et al., 2010), assessed only one variable: three used the variable Exam/Final grade (Fatemi et al., 2014;
Palocsay & Stevens, 2008; Woolley, 2015) all reporting neutral results; two used a pre-post-test measure, one showing positive (Fratto
et al., 2016) and the other neutral (Lee et al., 2010) results; one study evaluated homework completion (Johnston, 2004) and reported
neutral results; and one evaluated overall performance (Rodriguez & Smith, 2016) and also reported neutral results.
Regarding feedback, except for Woolley (2015), seven studies explicitly stated that the online platform provided some type of
feedback to students’ answers. Of these, except for two studies (Lee et al., 2010; Palocsay & Stevens, 2008), feedback provided was
solely in terms of correctness of the answer. Moreover, five studies assessed students’ perceptions towards the online homework format
(Fratto et al., 2016; Gaffney et al., 2010; Johnston, 2004; Rodriguez & Smith, 2016; Woolley, 2015). One out of five studies found that
students did not prefer one homework format over the other (Gaffney et al., 2010), while the other four studies reported that students
preferred the online format over the traditional one. Lastly, authors conclusions and recommendations about using the online format
for homework are as follows: four studies favored this format over the traditional one (Fratto et al., 2016; Gaffney et al., 2010;
Johnston, 2004; Rodriguez & Smith, 2016), three studies took no explicit position regarding which format they would recommend (Lee
et al., 2010; Palocsay & Stevens, 2008; Woolley, 2015), and the last study suggested combining traditional and online homework
formats (Fatemi et al., 2014).

4. Discussion

The current study aimed to conduct a systematic review of the research on the comparison of online homework and traditional
homework. The focus was twofold: i) learn the methodological features of the studies comparing online homework and traditional
homework, and ii) understand which format of homework has more benefits for student’s performance.
College students were the predominant population of the studies included in this review; with only two articles using alternative
distinct samples, namely fifth graders and high-school students. Science, Technology, Engineering, and Mathematics (STEM) and
Business were the areas represented in the articles reviewed, with a particular focus on mathematics (12 studies). This result may be
due to mathematics’ relevance for all areas of knowledge. Mathematical knowledge is undoubtedly needed in modern society and is
always present in technological development (Gonza �lez-Pienda, Nún
~ ez, Solano, Ros�
ario, Mour~ ao, Soares, & Valle, 2007). However, it
could also be the case that these types of questions and answers are a better fit for automatized correction by the computer than
open-ended answers.
Most of the papers included used convenience samples in a between-subjects design. This methodological option may have pre­
vailed due to participants’ availability (e.g., attending the researcher’s class). However, using convenience samples makes it difficult to
control influences that might interfere in the results other than the variable under study (online or traditional homework). For
example, the teachers’ approaches to teaching, the difficulty of the course perceived by students, or the students’ interest in the course
are among the variables likely to impact results and should be controlled for. The best way to control such influences is to randomize
the sample, but only one study in this review used this type of design (Jonsdottir et al., 2017). A clear line of scientific inquiry following
robust methodologies would have attained reliable data and concluded the efficacy of the distinct formats of presenting homework.
Twelve articles included a pre-post-test measure, but only three observed differences between groups; particularly, that online
homework was more effective than traditional homework (Arora et al., 2013; Fratto et al., 2016; Mendicino et al., 2009). Only one
article included a follow-up measure (Arora et al., 2013). The authors’ goal was to understand whether the gains obtained with the use

11
P. Magalh~
aes et al. Computers & Education 152 (2020) 103869

of online homework could be transferred to a different subject. Results showed that students in the online homework condition
presented higher academic results in the following course.
Typically, the studies included in the review were developed during a school year (two semesters). The first semester was used to
test the use of traditional homework and the other to test online homework. A semester (approximately 15 weeks) is a good duration of
time when researchers want to compare data between groups and learn differences and similarities. In contrast, the study by Woolley
(2015) lasted for four days. The limited duration of the interventions may prevent changes to occur and reliable conclusions regarding
results and the efficacy of the interventions.
Nineteen studies used platforms developed by a text book publisher. Only three investigations purposively developed a platform for
the study. Still, regardless of the origin of the online platform, the majority of the articles reported that their platform provided
feedback. The possibility of displaying feedback immediately after students complete their assignment is one of the advantages of
online homework (Davis & McDonald, 2016; Rodriguez & Smith, 2016). Instant feedback is unlikely to happen when homework is
done in the traditional format (Ros� ario et al., 2019). Typically, students in the latter situation must wait until the next class to receive
feedback (Bonham et al., 2001), and rarely is homework graded or personalized (e.g., individual comments providing suggestions on
how to improve) to fit students’ learning needs (e.g., Cunha et al., 2018). However, graded or commented homework is the feedback
most effective in promoting students’ achievement (Rosa �rio et al., 2015; Walberg, Paschal, & Weinstein, 1985), which may lead to
conclusions that teachers’ homework feedback may be less powerful than it could be (see Cunha et al., 2019; Ros� ario et al., 2015). The
feedback provided by a platform may help teachers overcome this limitation.
In the present systematic review, 24 of the 31 articles reported that the platforms selected included some type of feedback that
varied in terms of complexity. However, 8 of the 24 articles did not specify the type of feedback provided in their platforms (Demirci,
2007, 2010a, 2010b; Elmehdi et al., 2013; Gok, 2011; Lenz, 2010; Rodriguez & Smith, 2016; Stickles, 2017). The other 16 displayed
different types of feedback, such as stressing common errors, providing individualized coaching, providing extended consolidation
exercises, or checking correct or incorrect answers (Arora et al., 2013; Babaali & Gonzalez, 2015; Brewer & Becker, 2010; Burch & Kuo,
2010; Chow, 2014; Dufresne et al., 2002; Fatemi et al., 2014; Fratto et al., 2016; Fynewever, 2008; Gaffney et al., 2010; Hauk et al.,
2014; Johnston, 2004; Lee et al., 2010; Mathai & Olsen, 2013; Mendicino et al., 2009; Palocsay & Stevens, 2008).
Some authors have highlighted that one disadvantage of using online homework is the trial-and-error submissions, as it could
reinforce lazy behaviors (e.g., Bonham et al., 2001). In the present review, some studies have attempted to overcome this potential
limitation by adding a restriction to the number of attempts available to solve homework (Babaali & Gonzalez, 2015; Burch & Kuo,
2010; Davis & McDonald, 2016; Fatemi et al., 2014; Lenz, 2010; Woolley, 2015). Adding this feature might be beneficial for students,
because they can get immediate feedback on the number of tries and learn what they are doing wrong. Hopefully this strategy may
encourage them to continue working on the platform and complete exercises. Davis and McDonald (2016) followed a distinct strategy
to overcome the trial-and-error issue. They did not limit the number of attempts as a condition to complete exercises, but the score
students could achieve on a single homework was reduced as a function of the number of attempts used to find the correct answer. This
approach might discourage students that are trying hard to solve the exercise; in fact, they might avoid doing homework because they
may get a bad score (see also Kortemeyer, 2015).
Regarding the homework itself, overall, the information in the papers is scarce. The majority of the studies do not refer the purpose
underlying the assignment of homework, with the few that do highlighting the purpose of practicing concepts and skills (e.g., Hauk
et al., 2014; Mathai & Olsen, 2013). Also, despite the majority of the studies stating that the homework assignments were similar in the
online and traditional format, very few studies explicitly describe in detail the nature of the homework assignments. For the studies
that provide this information, the formats of the online answers were the following: multiple-choice; numerical answer; symbolic
(algebraic/mathematical) expression/answer; typing a word; short answer/essay; true or false; draw free-body diagrams or structures;
and upload a file (Brewer & Becker, 2010; Gok, 2011; Jonhston, 2004; Malik et al., 2014; Mathai & Olsen, 2013; Williams, 2010).
As many as half of the studies (15) were categorized as Neutral, meaning that both formats of delivering homework result in the
same outcome in terms of students’ performance. Only nine studies concluded that online homework, compared to traditional
homework, was more beneficial for students, and one study obtained the opposite result. Lastly, six studies reported mixed results
depending on the variable under study. Data is consistent with Mathai and Olsen (2013) who found that research on the effectiveness
of online homework in improving students’ performance has produced mixed results.
To deepen the understanding of these main results, particularly to learn if any conclusion can be drawn as a function of the course
domain, individual studies were grouped by course domain. The domains of Mathematics, Sciences, and Business were analyzed. The
majority of the studies in the domain College Business held neutral results, i.e. there are no differences for students’ performance as a
function of the homework format assigned. Conversely, the domains College Mathematics and College Sciences count with several studies
yielding positive results, i.e. students assigned to the online homework format had improved performance or higher results compared
with students assigned to the traditional homework format. It is unclear, however, the explanation for this difference between course
domains. When looking at the characteristics of the studies, the College Business studies provided limited information about the ty­
pology of the answers of the online homework; and the feedback provided by the online platform did not appear to have an instructive
purpose but more a correctional nature. In contrast, the majority of the studies in the College Mathematics and College Sciences domains
provided some detail about the format of the answers given to the homework activities, and the complexity of the feedback provided by
the platform was more elaborated and with instructional purposes besides the correctional function.
Additionally, half of the studies (15) included a measure to capture student’s perceptions and opinions regarding the online
homework format. In 13 of these studies, students reported to prefer and favor the online homework format over the traditional
homework format. Thus, overall, students are inclined towards the online homework format and highlight its many advantages.
Despite the overall null results, the majority of the authors of the individual studies recommend, and favor, the use of online

12
P. Magalh~
aes et al. Computers & Education 152 (2020) 103869

homework over the traditional homework format. Only five studies did not take an explicit position regarding this aspect (Gok, 2011;
Lee et al., 2010; Palocsay & Stevens, 2008; Williams, 2010; Woolley, 2015) and three suggest combining online and traditional
homework formats (Davis & McDonald, 2016; Fatemi et al., 2014; Mathai & Olsen, 2013). Interestingly, no study recommended the
traditional homework format over the online.

4.1. Educational and design implications

Overall, the articles reviewed stress advantages of adopting online homework. For example, by using the platform, faculty save
time preparing and checking homework, leaving them with more time to prepare classes and attend to students’ needs in the office.
Additionally, online homework is more fit for big classes, is environmentally friendly, and provides data regarding student’s per­
formance. Students may benefit as platforms may deliver exercises adjusted to their knowledge (more or less demanding exercises) and
provide immediate individualized feedback. Additionally, most publisher platforms possess a large set of exercises, creating the
possibility of endless generation of exercises, which could help students prepare for exams. By receiving instant feedback on an online
platform, time in class can be used to discuss particular aspects of the homework exercises assigned and check the problem-solving
strategies followed by their peers and the results found against their own output, thus improving their content knowledge. All
these online homework characteristics would help teachers manage the homework process, notwithstanding the possibility of
increased time spent by faculty answering questions on software related issues.
Regardless the benefits of online homework, it is important to note that when students do homework using the computer, they are
also likely to use it for purposes other than academics (e.g., surfing websites, playing games, listening to music) (Foehr, 2006; see also;
Calderwood, Ackerman, & Conklin, 2014), which may compromise students’ homework engagement (see Xu, 2015). For this reason,
assigning purposeful homework may help students understand its value and overcome the challenges to control volition (e.g., coping
with the technology-related distractions; see Xu, 2015). This aspect is particularly relevant considering that in the present review the
majority of the studies do not explicitly state the purpose of the homework assignments.
Interestingly, three studies (Davis & McDonald, 2016; Fatemi et al., 2014; Mathai & Olsen, 2013) suggest combining both tradi­
tional and online homework format as a function of the type of exercises, questions, and purposes of the homework. This suggestion
may provide some insights for developers designing educational computer systems. Particularly, in the future, these digital tools could
allow incorporating more complex and diversified type of exercises that foster the acquisition and practice of certain competences and
skills (e.g., draw structures). This aspect is particularly relevant considering that as many as half of the studies do not describe the
homework format by which the answers were entered in the online homework. Among those that describe the answers format, drawing
diagrams or graphs are seldom used, being multiple-choice and numeric answer the most common format. Furthermore, by designing
platforms that incorporate adaptive-responsive homework instead of the traditional-responsive homework (e.g., Richards-Babb,
Curtis, Ratcliff, Roy, & Mikalik, 2018) the platforms may favor students’ engagement and contribute to effective learning and aca­
demic success. Publishers and other relevant stakeholders designing educational computer systems could consider expanding their
platforms to domains other than STEM, but also to other grade levels. Considering that, worldwide, the majority of the students are in
grades other than college level, it is surprising that these tools have not yet emerged, or have not at least been extensively tested,
among elementary and secondary school level.

4.2. Limitations and future studies

In the current study, three databases were used in the search; however, and despite the novelty of this topic, a few papers comparing
online homework with traditional homework might have been left out. The present study did not control the quality of the papers
reviewed because the papers published fail to inform on important aspects of the investigation (e.g., number and education level of
participants; content domain of the homework assigned). Thus, future studies could consider having more robust study designs,
particularly in terms of controlling for the experimental conditions when comparing both formats of homework. For instance, three
articles had students in the online condition doing their homework assignments independently while the students in the traditional
condition did their assignments in group (Demirci, 2007; 2010a; 2010b). The author concluded that traditional homework condition
had better results in overall performance than online homework. In these cases, concluding for homework effectiveness would not be
precise because the conditions were not balanced. Doing homework in group or independently is a variable in itself with potential
impact on homework results (Demirci, 2010b). For more robust research, future research could also consider using randomized
samples of participants to strengthen the design and reduce bias, or using validated standardized measures to compare between
formats of homework.
Due to the purpose of this research, evaluation of students’ opinions was not an inclusion criterion during the selection of the
studies. Yet, 15 articles reviewed asked students their opinion about the use of online homework (see Table 4). Their opinion on online
homework, when compared to traditional homework, tended to be more positive. In general, students believe that online homework
has more benefits for their performance than traditional homework. This finding is consistent with data related to students’ general
understanding of homework benefits. For example, Woolley (2015) reported that students who believe that homework is helpful to get
a better test score are likely to complete more homework assignments than those who do not understand homework benefits (see also
Ros�ario et al., 2018). Thus, future reviews could consider further analyzing student’s opinions about the format of homework.
Future research could also consider exploring the use of online homework throughout schooling (e.g., elementary, middle, and high
school) to expand our understanding of the effects of online homework benefits for student’s progress. Particularly, researchers could
consider focusing their attention on elementary and middle school where students all over the world face important learning

13
P. Magalh~
aes et al. Computers & Education 152 (2020) 103869

challenges, particularly in mathematics and science (OECD, 2015). Focusing research on elementary and high school would be
especially important due to the massive use of the internet and technological devices by young students (OECD, 2015).
Lastly, all the papers included in the sample had a focus on STEM, perhaps because it is easier to develop exercises and provide
feedback to be delivered by online homework systems. However, it would be interesting to explore other domains, such as languages or
social sciences to compare results on the efficacy of these platforms.

5. Conclusions

The current study aimed to systematize the methodological features of the studies comparing online and traditional homework
formats and understand which of the two formats benefits most students’ performance. We believe that findings are likely to help
practitioners, researchers, and educational systems developers consolidate knowledge in the literature and learn the research and
design gaps that could be addressed in the future.
Regarding the first research question, results showed that: i) the majority of the studies included in this review were published from
2010 onwards; ii) most studies included college students and presented a wide range of number of participants, duration of inter­
vention, and course domain; iii) the majority of the studies used a convenience sample and a between-subject study design; iv) studies
used different approaches to implement the online homework; and v) for the majority of the studies the outcome measures were
assessed by non-standardized instruments. This variety makes the task of answering the second research question challenging. Results
showed that as many as half of the studies reported neutral results. That is, no differences were found between online and traditional
homework regarding students’ performance. Thus, one can conclude that both online and traditional format are, at least, equally
effective. What is more, results seem to differ as a function of the course domain. Particularly, in the College Business course domain, the
majority of the studies reported neutral results. Whereas in the other course domains there were several studies showing positive and
mixed results. As previously mentioned, descriptions regarding the characteristics of the study (e.g., typology of the answers of the
online homework) and the type of feedback (i.e. correctional nature rather than an instructive one) provided by the College Business
studies were limited. Thus, there is the possibility that these studies have some methodological shortcomings and that further more
robust research (e.g., RCTs) will help shed some light over this question. Nevertheless, there is also the possibility that the online
homework format may not be equally effective for different course domains, with more research being needed to address this
particular aspect. Lastly, about half of the studies assessed students’ perceptions and opinions about the online format and, except for
two studies, students were favorable towards the online homework format and preferred it over the traditional format. Overall, despite
the null results, the majority of the studies in this sample concluded in favor of using the online format for delivering homework, while
none favored the traditional format of homework.
Overall, there is no unanimous answer regarding which format of homework favors students the most in terms of school
achievement. However, when examining indicators of a distinct nature, i.e. students’ perceptions and conclusions of the authors of the
current sample of studies, there is a consensus that online homework contributes more to students’ performance and favors more
students’ engagement than traditional homework. Despite the need to deepen the knowledge and conduct more robust research
comparing the effect of the homework format on students’ achievement, it is unanimous among participants and authors of this sample
of studies that online homework engages students to perform the tasks. This outcome cannot be disregarded as school engagement is a
strong predictor of academic achievement (Finn & Zimmer, 2012). Thus, if students consistently perceive that online homework
promotes deeper learning and more academic success, this by itself is a strong argument favoring the adoption of this homework
format.
This systematic review highlights the need to investigate this topic in lower school grades. Considering the strong investment made
to equip elementary and secondary schools with technology in recent years and the massive use of the internet and technological
devices by young students, future research could explore online homework at elementary, middle, and high school levels. These results
may prove useful to increase the efficacy of homework as a tool to promote student engagement and performance, particularly at
middle school level where the relationship between traditional homework and performance is weaker than that found in the other
school levels (Fan et al., 2017).
Regarding educational implications, results showed that online homework: i) allows to assign exercises adjusted to students’
knowledge, with the possibility of randomization; ii) is a good solution for big classes, such as at college level; and iii) provides im­
mediate feedback to teachers and students about homework performance and achievement, which is a great challenge in traditional
homework format (Cunha et al., 2018; Rosa �rio et al., 2019). Lastly, results led us to reflect that a hybrid approach, adjusted for the
types of exercises and questions, may be beneficial for students.

6. Declarations of interest

None.

Funding

This study was conducted at Psychology Research Centre (UID/PSI/01662/2013), University of Minho, and supported by the
Portuguese Foundation for Science and Technology and the Portuguese Ministry of Science, Technology and Higher Education through
national funds and co-financed by FEDER through COMPETE2020 under the PT2020 Partnership Agreement (POCI-01-0145-FEDER-
007653) (POCI-01-0145-FEDER-028302).

14
P. Magalh~
aes et al. Computers & Education 152 (2020) 103869

CRediT authorship contribution statement

Paula Magalha ~es: Conceptualization, Methodology, Formal analysis, Investigation, Data curation, Writing - original draft, Writing
- review & editing, Visualization. Diogo Ferreira: Formal analysis, Investigation, Data curation, Writing - original draft, Visualization.
Jennifer Cunha: Formal analysis, Investigation, Writing - review & editing, Visualization. Pedro Rosa � rio: Conceptualization,
Methodology, Writing - review & editing, Supervision, Funding acquisition.

Acknowledgements

Authors would like to thank Connor Holmes for the English editing of the manuscript.

References*

*Arora, M. L., Rho, Y. J., & Masson, C. (2013). Longitudinal study of online statistics homework as a method to improve learning. Journal of STEM Education, 14(1),
36–43.
*Babaali, P., & Gonzalez, L. (2015). A quantitative analysis of the relationship between an online homework system and student achievement in pre-calculus.
International Journal of Mathematical Education in Science & Technology, 46(5), 687–699. https://doi.org/10.1080/0020739x.2014.997318.
Bonham, S., Beichner, R., & Deardorff, D. (2001). Online homework: Does it make a difference? The Physics Teacher, 39, 293–296. https://doi.org/10.1119/
1.1375468.
*Bonham, S. W., Deardorff, D. L., & Beichner, R. J. (2003). Comparison of student performance using web and paper-based homework in college-level physics. Journal
of Research in Science Teaching, 40(10), 1050–1071. https://doi.org/10.1002/tea.10120.
*Brewer, D. S., & Becker, K. (2010). Online homework effectiveness for underprepared and repeating college algebra students. Journal of Computers in Mathematics and
Science Teaching, 29(4), 353–371.
Buijs, M., & Admiraal, W. (2013). Homework assignments to enhance student engagement in secondary education. European Journal of Psychology of Education, 28(3),
767–779. https://doi.org/10.1007/s10212-012-0139-0.
*Burch, K. J., & Kuo, Y. (2010). Traditional vs. online homework in college algebra. Mathematics and Computer Education, 44(1), 53–63.
Calderwood, C., Ackerman, P. L., & Conklin, E. M. (2014). What else do college students “do” while studying? An investigation of multitasking. Computers & Education,
75, 19–29. https://doi.org/10.1016/j.compedu.2014.02.004.
Chang, C. B., Wall, D., Tare, M., Golonka, E., & Vatz, K. (2014). Relationships of attitudes toward homework and time spent on homework to course outcomes: The
case of foreign language learning. Journal of Educational Psychology, 106(4), 1049–1065. https://doi.org/10.1037/a0036497.
Chan, T., Roschelle, J., Hsi, S., K., Sharples, M., Brown, T., et al. (2006). One-to-one technology-enhanced learning: An opportunity for global research collaboration.
Research and Practice in Technology Enhanced Learning, 1(1), 3–29. https://doi.org/10.1142/s1793206806000032.
*Chow, A. F. (2014). Online homework impact in undergraduate mathematics and business statistics courses. Educational Studies, 41(3), 244–248. https://doi.org/
10.1080/03055698.2014.961902.
Cooper, H. (2001). The battle over homework (2nd ed.). Corwin Press.
Cooper, H., Lindsay, J. J., Nye, B., & Greathouse, S. (1998). Relationships among attitudes about homework, amount of homework assigned and completed, and
student achievement. Journal of Educational Psychology, 90(1), 70–83. https://doi.org/10.1037/0022-0663.90.1.70.
Cooper, H., Robinson, J. C., & Patall, E. A. (2006). Does homework improve academic achievement? A synthesis of research, 1987–2003. Review of Educational
Research, 76(1), 1–62. https://doi.org/10.3102/00346543076001001.
Cunha, J., Ros� ario, P., Nú~
nez, J. C., Nunes, A. R., Moreira, T., & Nunes, T. (2018). “Homework feedback is…”: Elementary and middle school teachers’ conceptions of
homework feedback. Frontiers in Psychology, 9, 1–20. https://doi.org/10.3389/fpsyg.2018.00032.
Cunha, J., Ros� ario, P., Nú~
nez, J. C., Vallejo, G., Martins, J., & H€
ogemann, J. (2019). Does teacher homework feedback matter to 6th graders’ school engagement?: a mixed
methods study. Metacognition and Learning, 14, 89–129. https://doi.org/10.1007/s11409-019-09200-z.
*Davis, J. L., & McDonald, T. N. (2016). Online, handwritten or hybrid homework: What’s best for our students in the long run? Journal of Online Engineering
Education, 7(1).
*Demirci, N. (2007). University students perceptions of web-based vs. paper-based homework in a general physics course. Eurasia Journal of Mathematics, Science and
Technology Education, 3(1), 29–34. https://doi.org/10.12973/ejmste/75371.
*Demirci, N. (2010a). Web-based vs. paper-based homework to evaluate students’ performance in introductory physics courses and students’ perceptions: Two years
experience. International Journal on E-Learning, 9(1), 27–49.
*Demirci, N. (2010b). The effect of web-based homework on university students’ physics achievement. The Turkish Online Journal of Educational Technology, 9(4),
156–161.
Dettmers, S., Trautwein, U., Lüdtke, O., Goetz, T., Frenzel, A. C., & Pekrun, R. (2011). Students’ emotions during homework in mathematics: Testing a theoretical
model of antecedents and achievement outcomes. Contemporary Educational Psychology, 36(1), 25–35. https://doi.org/10.1016/j.cedpsych.2010.10.001.
Dettmers, S., Trautwein, U., Lüdtke, O., Kunter, M., & Baumert, J. (2010). Homework works if homework quality is high: Using multilevel modeling to predict the
development of achievement in mathematics. Journal of Educational Psychology, 102(2), 467–482. https://doi.org/10.1037/a0018453.
Dillard-Eggers, J., Wooten, T., Childs, B., & Coker, J. (2008). Evidence on the effectiveness of on-line homework. College Teaching Methods & Styles Journal, 4(5), 9–16.
https://doi.org/10.19030/ctms.v4i5.5548.
*Dodson, J. R. (2014). The impact of online homework on class productivity. Science Education International, 25(4), 354–371.
*Dufresne, R., Mestre, J., Hart, D. M., & Rath, K. A. (2002). The effect of web-based homework on test performance in large enrollment introductory physics courses.
Journal of Computers in Mathematics and Science Teaching, 21(3), 229–249.
Eichler, J. F., & Peeples, J. (2013). Online homework put to the test: A report on the impact of two online learning systems on student performance in general
chemistry. Journal of Chemical Chemistry, 90, 1137–1143. dx.doi.org/10.1021/ed3006264.
Elawar, M. C., & Corno, L. (1985). A factorial experiment in teachers’ written feedback on student homework: Changing teacher behavior a little rather than a lot.
Journal of Educational Psychology, 77, 162–173.
*Elmehdi, H. M., Ibrahim, A. M., & Haba, U. A. (2013). An evaluation of web-based homework (WBH) delivery systems: University of sharjah experience. International
Journal of Emerging Technologies in Learning (iJET), 8(4), 57–62. https://doi.org/10.3991/ijet.v8i4.2966.
Epstein, J. L., & Van Voorhis, F. L. (2001). More than ten minutes: Teachers’ roles in designing homework. Educational Psychologist, 36(3), 181–193. https://doi.org/
10.1207/S15326985EP3603_4.
Fan, H., Xu, J., Cai, Z., He, J., & Fan, X. (2017). Homework and students’ achievement in math and science: A 30-year meta-analysis, 1986–2015. Educational Research
Review, 20, 35–54. https://doi.org/10.1016/j.edurev.2016.11.003.
*Fatemi, D., Marquis, L., & Wasan, S. (2014). Student performance in intermediate accounting: A comparison of the effectiveness of online and manual homework
assignments. The Accounting Educators’ Journal, 1–19.

*
This reference was a result of the literature search for this systematic review.

15
P. Magalh~
aes et al. Computers & Education 152 (2020) 103869

Fern�andez-Alonso, R., Alvarez-Díaz,

� M., Su�arez-Alvarez,
� J., & Mu~ niz, J. (2017). Students’ achievement and homework assignment strategies. Frontiers in Psychology, 8,
1–11. https://doi.org/10.3389/fpsyg.2017.00286.
Finn, J. D., & Zimmer, K. S. (2012). Student engagement: What is it? Why does it matter? In S. L. Christenson, A. L. Reschly, & C. Wylie (Eds.), Handbook of research on
student engagement (pp. 97–131). New York, NY: Springer. https://doi.org/10.1007/978-1-4614-2018-7_5.
Foehr, U. G. (2006). Media multitasking among American youth: Prevalence, predictors, and pairings. Menlo Park, CA: Kaiser Family Foundation.
Foyle, H., Lyman, L., Tompkins, L., Perne, S., & Foyle, D. (1990). Homework and cooperative learning: A classroom field experiment. Emporia, KS: Emporia State
University.
*Fratto, V., Sava, M. G., & Krivacek, G. J. (2016). The impact of an online homework management system on student performance and course satisfaction in
introductory financial accounting. International Journal of Information and Communication Technology Education, 12(3), 76–87. https://doi.org/10.4018/
ijicte.2016070107.
Fulano, C., Cunha, J., Nú~ nez, J. C., Pereira, B., & Ros�
ario, P. (2018). Mozambican adolescents’ perspectives on the academic procrastination process. School Psychology
International, 39(2), 196–213. https://doi.org/10.1177/0143034318760115.
*Fynewever, H. (2008). A comparison of the effectiveness of web-based and paper-based homework for general chemistry. The Chemical Educator, 13, 264–269.
https://doi.org/10.1333/s00897082142a.
*Gaffney, M. A., Ryan, D., & Wurst, C. (2010). Do online homework systems improve student performance? Advances in Accounting Education, 49–68. https://doi.org/
10.1108/s1085-4622(2010)0000011005.
Galloway, M., Conner, J., & Pope, D. (2013). Nonacademic effects of homework in privileged, high-performing high schools. The Journal of Experimental Education, 81
(4), 490–510. https://doi.org/10.1080/00220973.2012.745469.
*Gok, T. (2011). Comparison of student performance using web- and paper- based homework in large enrollment introductory physics courses. International Journal of
the Physical Sciences, 6(15), 3778–3784. https://doi.org/10.5897/IJPS11.217.
G€ollner, R., Damian, R. I., Rose, N., Spengler, M., Trautwein, U., Nagengast, B., et al. (2017). Is doing your homework associated with becoming more conscientious?
Journal of Research in Personality, 71, 1–12. https://doi.org/10.1016/j.jrp.2017.08.007.
Gonz� alez-Pienda, J. A., Nú~ nez, J. C., Solano, P., Ros�ario, P., Mour~ ao, R., Soares, S., & Valle, A. (2007). Atitudes Face A
� Matem�atica E Rendimento Escolar No Sistema
Educativo Espanhol. Psicologia: Teoria, Investigaç~ ao e Pr� atica, 1, 151–160.
Hagger, M., Sultan, S., Hardcastle, S., & Chatzisarantis, N. (2015). Perceived autonomy support and autonomous motivation toward mathematics activities in
educational and out-of-school contexts is related to mathematics homework behavior and attainment. Contemporary Educational Psychology, 41, 111–123. https://
doi.org/10.1016/j.cedpsych.2014.12.002.
Halcrow, C., & Dunnigan, G. (2012). Online homework in calculus I: Friend or foe? Primus, 22(8), 664–682. https://doi.org/10.1080/10511970.2012.694015.
*Hauk, S., Powers, R. A., & Segalla, A. (2014). A comparison of web-based and paper-and-pencil homework on student performance in college algebra. Primus, 25(1),
61–79. https://doi.org/10.1080/10511970.2014.906006.
Ismail, M., Mokhtar, W. Z., Nasir, N. N., Rashid, N. R., & Ariffin, A. K. (2014). The development of a web-based homework system (WBH) via TCExam. Mediterranean
Journal of Social Sciences. https://doi.org/10.5901/mjss.2014.v5n15p249.
*Johnston, T. C. (2004). Online homework assessments: Benefits and drawbacks to students. Academy of Educational Leadership Journal, 8(3), 29–40.
*Jonsdottir, A. H., Bjornsdottir, A., & Stefansson, G. (2017). Difference in learning among students doing pen-and-paper homework compared to web-based
homework in an introductory statistics course. Journal of Statistics Education, 25(1), 12–20. https://doi.org/10.1080/10691898.2017.1291289.
Kaur, B. (2011). Mathematics homework: A study of three grade eight classrooms in Singapore. International Journal of Science and Mathematics Education, 9(1),
187–206. https://doi.org/10.1007/s10763-010-9237-0.
Keith, T. Z. (1982). Time spent on homework and high school grades: A large-sample path analysis. Journal of Educational Psychology, 74(2), 248–253.
Kortemeyer, G. (2015). An empirical study of the effect of granting multiple tries for online homework. American Journal of Physics, 83(7), 646–653. https://doi.org/
10.1119/1.4922256.
*Lee, W., Courtney, R. H., & Balassi, S. J. (2010). Do online homework tools improve student results in principles of microeconomics courses? The American Economic
Review, 100(2), 283–286. https://doi.org/10.1257/aer.100.2.283.
*Lenz, L. (2010). The effect of a Web-Based Homework System on student outcomes in a first-year mathematics course. Journal of Computers in Mathematics and Science
Teaching, 29(3), 233–246.
Lim, C.-P., Zhao, Y., Tondeur, J., Chai, C.-S., & Tsai, C.-C. (2013). Bridging the gap: Technology trends and use of technology in schools. Educational Technology &
Society, 16(2), 59–68.
Lunsford, M. L., & Pendergrass, M. (2016). Making online homework work. Primus, 26(6), 531–544. https://doi.org/10.1080/10511970.2015.1110219.
*Malik, K., Martinez, N., Romero, J., Schubel, S., & Janowicz, P. A. (2014). Mixed-methods study of online and written organic chemistry homework. Journal of
Chemical Education, 91(11), 1804–1809. https://doi.org/10.1021/ed400798t.
*Mathai, E., & Olsen, D. (2013). Studying the effectiveness of online homework for different skill levels in a college algebra course. Primus, 23(8), 671–682. https://
doi.org/10.1080/10511970.2013.782479.
*Mendicino, M., Razzaq, L., & Heffernan, N. T. (2009). A comparison of traditional homework to computer-supported homework. Journal of Research on Technology in
Education, 41(3), 331–359. https://doi.org/10.1080/15391523.2009.10782534.
OECD. (2014). PISA 2012 Results in focus: What 15-year-olds know and what they can do with what they know. PISA. Paris: OECD Publishing.
OECD. (2015). Students, computers and learning: Making the connection. PISA: OECD Publishing. https://doi.org/10.1787/9789264239555-en.
Higgins, J. P. T., & Wells, G. A. (2011). Cochrane handbook for systematic reviews of interventions.
Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009). Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Annals of
internal medicine, 151(4), 264–269.
Nú~nez, J. C., Su�arez, N., Cerezo, R., Gonz� alez-Pienda, J., Ros� ario, P., Mour~ao, R., & Valle, A. (2013). Homework and academic achievement across Spanish
Compulsory Education. Educational Psychology, 35(6), 726–746. https://doi.org/10.1080/01443410.2013.817537.
Nú~nez, J. C., Su�arez, N., Ros� ario, P., Vallejo, G., Cerezo, R., & Valle, A. (2014). Teachers’ feedback on homework, homework-related behaviors, and academic
achievement. The Journal of Educational Research, 108(3), 204–216. https://doi.org/10.1080/00220671.2013.878298.
*Palocsay, S. W., & Stevens, S. P. (2008). A Study of the effectiveness of web-based homework in teaching undergraduate business statistics. Decision Sciences Journal
of Innovative Education, 6(2), 213–232. https://doi.org/10.1111/j.1540-4609.2008.00167.x.
Parker, L. L., & Loudon, G. M. (2013). Case study using online homework in undergraduate organic chemistry: Results and student attitudes. Journal of Chemical
Education, 90, 37–44. dx.doi.org/10.1021/ed300270t.
Proctor, M. D., & Marks, Y. (2013). A survey of exemplar teachers’ perceptions, use, and access of computer-based games and technology for classroom instruction.
Computers & Education, 62, 171–180. https://doi.org/10.1016/j.compedu.2012.10.022.
Richards-Babb, M., Curtis, R., Ratcliff, B., Roy, A., & Mikalik, T. (2018). General chemistry student attitude and success with use of online homework: Traditional-
responsive versus adaptive-responsive. Journal of Chemical Education, 95, 691–699. https://doi.org/10.1021/acs.jchemed.7b00829.
*Rodriguez, P., & Smith, L. M. (2016). Using an online homework management system in tax accounting: Does it advance learning? International Journal of Learning
Technology, 11(1), 44–65. https://doi.org/10.1504/ijlt.2016.076062.
Ros�ario, P., Cunha, J., Nunes, A. R., Moreira, T., Nú~ nez, J. C., & Xu, J. (2019). Did you do your homework?” Mathematics teachers’ homework follow-up practices at
middle school level. Psychology in the Schools, 56(1), 92–108. https://doi.org/10.1002/pits.22198.
Ros�ario, P., Cunha, J., Nunes, T., Nunes, A. R., Moreira, T., & Nú~ nez, J. C. (2019). Homework should be… but we do not live in an ideal world”: Mathematics teachers’
perspectives on quality homework and on homework assigned in elementary and middle schools. Frontiers in Psychology, 10, 1–15. https://doi.org/10.3389/
fpsyg.2019.00224.
Ros�ario, P., Mour~ ao, R., Baldaque, M., Nunes, T., Nú~ nez, J., Gonz� alez-Pienda, J. A., & Valle, A. (2009). Tareas para casa, autorregulaci�
on del aprendizaje y
rendimiento en matem� aticas. Revista de Psicodid�actica, 14(2), 179–192.

16
P. Magalh~
aes et al. Computers & Education 152 (2020) 103869

Ros�ario, P., Nú~nez, J. C., Vallejo, G., Cunha, J., Nunes, T., Su�
arez, N., & Moreira, T. (2015). Teachers’ homework follow-up practices and students’ EFL performance: A
randomized group design. Frontiers in Psychology, 6, 1–11. https://doi.org/10.3389/fpsyg.2015.01528.
Ros�ario, P., Nú~ nez, J. C., Vallejo, G., Nunes, T., Cunha, J., Fuentes, S., & Valle, A. (2018). Homework purposes, homework behaviors, and academic achievement.
Examining the mediating role of students’ perceived homework quality. Contemporary Educational Psychology, 53, 168–180. https://doi.org/10.1016/j.
cedpsych.2018.04.001.
Shute, V. J. (2008). Focus on formative feedback. Review of Educational Research, 78(1), 153–189. https://doi.org/10.3102/0034654307313795.
Smith, B. A., & McCoy, L. P. (2011). Using homework to improve students’ abilities to self-regulate. In Studies in teaching – 2011 research digest (pp. 121–126).
*Stickles, P. R. (2017). The implementation of online homework in college algebra and its results. Illinois Mathematics Teacher, 1–10.
Trautwein, U. (2007). The homework-achievement relation reconsidered: Differentiating homework time, homework frequency, and homework effort. Learning and
Instruction, 17(3), 372–388. https://doi.org/10.1016/j.learninstruc.2007.02.009.
Trautwein, U., & Lüdtke, O. (2007). Students’ self-reported effort and time on homework in six school subjects: Between-students differences and within-student
variation. Journal of Educational Psychology, 99(2), 432–444. https://doi.org/10.1037/0022-0663.99.2.432.
Trautwein, U., Niggli, A., Schnyder, I., & Lüdtke, O. (2009). Between-teacher differences in homework assignments and the development of students’ homework
effort, homework emotions, and achievement. Journal of Educational Psychology, 101, 176–189. https://doi.org/10.1037/0022-0663.101.1.176.
Walberg, H. J., & Paik, S. J. (2000). Effective educational practices. Brussels: International Academy of Education & International Bureau of Education.
Walberg, H. J., Paschal, R. A., & Weinstein, T. (1985). Homework’s powerful effects on learning. Educational Leadership, 42, 76–79.
*Williams, A. (2010). Online homework vs. traditional homework: Statistics anxiety and self-efficacy in an educational statistics course. Technology Innovations in
Statistics Education, 6(1).
*Woolley, D. J. (2015). Which helps accounting students learn more: Traditional homework, online homework, or clickers? Academy of Educational Leadership Journal,
19(3), 337–343.
Xu, J. (2015). Investigating factors that influence conventional distraction and tech-related distraction in math homework. Computers & Education, 81, 304–314.
https://doi.org/10.1016/j.compedu.2014.10.024.

17