Performance-based accountability: exploring Ghanaian teachers perception of the influence of large-scale testing on teaching and learning

ABSTRACT

Undergirded by “economic theory of the principal–agent problem”, the study investigated secondary school teachers’ perception of the influence of large-scale testing accountability on teaching and learning. Cross-sectional survey design was used. Simple random sampling was also employed to select 200 teachers for this study. The exploratory factor analysis results showed that although large-scale testing accountability serves as a motivational factor for teachers to be more productive, it puts pressure on teachers and students. Teachers reported teaching to the test (i.e. washback) and spending a lot of instructional time preparing students for the test. Teachers suggested that large-scale testing results should not be seen as the sole basis for teacher effectiveness. Rather, they indicated that large-scale testing results should be used to foster equitable distribution of educational resources and support teachers’ professional development to enhance teaching and learning. Implications for policy and practice are discussed.

KEYWORDS:

• Ghanian teachers
• large scale testing
• secondary school
• accountability

Introduction

Accountability could be understood as either being held accountable or giving an account (Lingard, Sellar, & Lewis, 2017). Depending on the phenomenon, accountability could be seen as either answerability or responsibility (Gregory, 2003). Answerability accentuates that people or some group of individuals should be held accountable for either or not achieving a required and agreed upon goals. Responsibility highlights the conceptions that people or group of individuals must be held accountable for upholding norms or rules that are mostly unclear and very complicated to specify. Researchers have argued that tensions between answerability and responsibility have shaped accountability policies in most countries (Olssen & Peters, 2005; Suspitsyna, 2010). In education, accountability has garnered global attention largely because of how it has historically been perceived as a one-way responsibility of educators to demonstrate to governments that outcomes have been realised or that policies have been either or not effectively implemented (Edmonton, 2005). Contemporary, educational scholars perceive accountability within education as not static, but rather a complex and frequently changing social process (Lingard et al., 2017). In lieu of the constantly evolving notion regarding accountability, there are varying conceptualisation and approaches to accountability, which are concurrently at play and in competition for recognition within school systems. Among these accountability conceptualisations are but not limited to bureaucratic or performance-based accountability (Darling-Hammond, 1989; Firestone & Shipps, 2007), intelligent accountability (O’Neill, 2013; Sahlberg, 2010), genuine accountability (Darling-Hammond, Wilhoit, & Pittenger, 2014) and professional accountability (Darling-Hammond, 1989; Firestone & Shipps, 2007).

Despite different operationalisation and conceptualisation of accountability in education (Gilbert, 2011), it basically narrows down to key actors and agents within the school system (i.e. schools, headteachers, students and teachers) who are being held accountable for student performance, especially in a large-scale testing (Brill, Grayson, Kuhn, & O’Donnell, 2018). For this study, we focused on bureaucratic accountability that over a decade has been conceptualised as performance-based accountability (Firestone & Shipps, 2007). Bureaucratic accountability has historically been understood as hierarchically structured in a bureaucratic manner. In this approach to accountability “priorities determined at the superior level are followed, and supervisory control is exercised intensively with a clear understanding for the need to follow orders” (Stone, 1989, p. 6). Contemporary, bureaucratic or performance-based accountability approach largely refers to the use of data from large-scale student testing as a mechanism or tool to hold school systems accountable (Popham, 1999). Although extant literature has demonstrated that one major aim of large-scale testing is to support student success by checking those in charge of educating students (Decker & Bolt, 2008; Kellaghan, Greaney, & Murray, 2009; Klinger, DeLuca, & Miller, 2008), unfortunately large-scale testing has come to be a tool for accountability across the globe, and testing has shifted from an instrument for making significant decisions about learners to a tool for holding schools and educational systems accountable (Brill et al., 2018; Chung & Chea, 2016; Hargreaves, 2020; Kirkpatrick & Zang, 2011; Liu, Luo, & Tang, 2021; Rahman, Seraj, Hasan, Namaziandost, & Tilwani, 2021).

In sub-Saharan Africa, large-scale testing has been found to be imperative to all educational stakeholders and policy makers (Browne, 2016). Besides evaluating the quality of schools and educational systems, large-scale testing also determines the future career of students (Browne, 2016). Large-scale testing has been used in Sub-Saharan Africa as a gate keeping tool because of the limited and scarce resources. Accountability issues, in response to political, social, and economic pressures, have in over the years gained recognition and importance in government administrations in many countries. This accountability system varies, unfolding with different speeds and impact (Hopmann 2007; Kellaghan et al., 2009). Increasingly political accountability demands have led to proliferation of testing requirements, which put teachers and students under constant monitoring to make sure learning is effectively and successfully happening (Barksdale-Ladd & Thomas, 2000; Hargreaves, 2020; Hoffman, Assaf, & Paris, 2001; Kellaghan et al., 2009; Liu et al., 2021; Popham, 2003; Rahman et al., 2021).

Large scale testing: advocates and critics

Due to the diverse usage of large-scale testing in schools and educational systems, it has become one of the highly contentious topics in educational jurisdictions around the globe. Activists of large-scale testing are of the view that these tests put the need pressure on educators to work effectively and efficiently. Advocates of large-scale testinguphold that it offers students with knowledge of their own learning and also teachers with deeper understanding of students’ progress (Carnoy & Loeb, 2002; Hamilton & Gonzales, 2003; Roderick, Jacob & Bryk, 2002). Researchers also argued that high-stake standardised exams are strongly linked to improvements in an average performance in international standardised assessments (Bergbauer, Hanushek, & Woessmann, 2018; Yang, Luo, Vadillo, Yu, & Shanks, 2021). Similarly, other researchers assert that when teachers and schools are held responsible for student performance, performing schools and teachers are identified and rewarded and underperforming schools can be supported with extra resources to ensure equitable distribution of resources (Elbousty, 2009; Jones & Egley, 2007; Marchant & Paulson, 2005).

Advocates further argue that performance-based accountability programmes are to reduce the achievement disparities between students from high socio-economic status and those from low socio-economic status. Accountability systems are based on the premise that if schools are held accountable for scores on large-scale tests, student attainment will increase, mainly in schools that serve a large number of students from poor homes. Moreover, awards tied to student achievement are supposed to encourage and stimulate administrators, teachers, and students to improve educational practices (i.e. instructional practices and learning strategies) in order to achieve positive results. The testing results are aimed at identifying learners and schools in need of remedy, thus permitting a more effective use of resources (Hamilton & Koretz, 2002).

In contrast, critics of large-scale tests argue that large-scale testing accountability augment students’ probability of low achievement and hold teachers with unequal and inequitable resources to the same standards, narrow and garble the curriculum, and coagulate social class and ethnic discrepancies (Berliner, 2011; Bracey, 2000; Burger & Krueger, 2003; Mika, 2005; Nichols & Berliner, 2008; Rahman et al., 2021). Additionally, they contend that high stakes testing accountability transmits the information that the central aim of learning is to get good results (Berliner, 2011; Klenowski, 2011; Klinger & Rogers, 2011; Nichols & Berliner, 2008). Hargreaves (2020) maintained that high stake testing has led to “teaching to the test, cultural bias, avoidance of innovation, dilemmas of whether to include highly vulnerable students in the testing process or not, and emotional ill-being among students and teachers” (P. 414). Similarly, extant literature shows that high-stake testing reduces significant amount of time spent in deep-rooted learning and vital creative skills including self-monitoring skills, organisation of knowledge, problem-solving etc (Ananda & Ashadi, 2021; Berliner, 2011; Klenowski, 2011; Klinger & Rogers, 2011; Zhao, 2015). Irrespective of others perception, the importance and efficiency of high-stake testing, it is unquestionable that these testing programmes have had substantial adverse impacts on students, teachers, and school administrators. For example, countries develop large-scale assessments that hold principals and students accountable. In these systems, students do not graduate unless they pass an exit exam and principals do not get raises unless student achievement improves to a predetermined level. Principals feel the pressure to improve student achievement and fear losing their contracts if they fail to make academic gains (Berliner, 2011; Kellaghan et al., 2009; Liu et al., 2021; Prytula, Noonan, & Hellsten, 2013).

Moreover, critics of high-stakes accountability do not believe that large-scale testing when used for making high decisions raise the achievement disparities between privileged and underprivileged students (von Zastrow & Janc, 2004). Rather, they contend that the pressure for schools to meet adequate yearly progress and prevent punishments urges teachers to concentrate only on the content and skills assessed on the tests. This political accountability leads to narrowing of the curriculum and thus restricts the experiences of students, notably students who are in schools that serve larger populations of underprivileged students (Thompson, 2001). Additionally, the pressure to increase large-scale test results might adversely affect student placement decisions (Baidoo-Anu, 2022; Baidoo-Anu, Gyamerah, & Chanimbe, 2022), such as class retention and special education detection. Critics also claim that test-based accountability policies make it complicated for schools to hire and retain qualified teachers (Rotberg, Bernstein, & Ritter, 2001).

The Ghanaian context

Various countries, including Ghana have made large-scale testing the centrepiece of their school reform (Anane, 2015). In lieu of this, large-scale testing is prevalent in the educational system in Ghana. Students at all levels of education from kindergarten through to junior high school are required to take these tests. Irrespective of the grade level, teacher and national-mandated tests are very competitive and results are used for high-stake decisions: they determine students’ progress to the next grade level (Baidoo-Anu, 2022; Baidoo-Anu et al., 2022). At present, national-mandated tests such as the Basic Education Certificate Examination (BECE) (i.e. school-leaving examination all Grade 9 students must write and pass before progressing to secondary school) and West African Senior School Certificate Examination (WASSCE) (i.e. qualifying exams that all Grade 12 students write and pass before progressing to have post-secondary education) have become high-stakes (Amoako, 2019). Outcomes of these national mandated tests are also used for measuring the quality and effectiveness of school, and teachers. Although several studies have investigated the influence of large-scale testing accountability on curriculum and learning at the basic level (Amoako, 2019), there are limited studies on the influence of large-scale testing accountability on senior high-school teachers teaching and students learning. As a preliminary investigation, this study unpacks secondary school teachers’ perception of the influence of large-scale testing accountability on teaching and learning and suggests ways that data from large-scale testing could be used to move learning forward. Based on the purpose of the study, two research questions were formulated to guide the study.

1. What are senior high-school teachers’ perceptions of the influences of large-scale testing accountability on teachers teaching and learning?

2. What do teachers suggest are effective ways to use data from large-scale testing?

Theoretical framework

Based on the overarching objective of the study, the economic theory of the principal–agent problem was employed to provide a theoretical lens for the study. Economic theory of the principal–agent problem was proposed by Grossman and Hart in 1983. This theory has been used extensively to provide a justification for the proliferation of the usage of data-driven accountability in schools and educational systems (Figlio & Loeb, 2011; Rothstein, Jacobsen, & Wilder, 2008). Principal–agent problem happens when one person or group who is the agent/s (such as schools, teachers) acts on behalf of another person or group who is the principal/s (such as parents, governments). According to the theory, suboptimal outcomes occur when there is misalignment in the aims and incentives of the principals (i.e. parents and governments) and the agents (i.e. schools and teachers). Theory of the principal–agent problem was deemed appropriate for this study since in bureaucratic or performance-based approaches to accountability “priorities determined at the superior level are followed, and supervisory control is exercised intensively with a clear understanding for the need to follow orders” (Stone, 1989, p. 6). The theory provides understanding of how principal (i.e. governments) uses data-driven accountability to force agents (i.e. schools, teachers and headteachers) to align their goals and objectives to the goals and objectives of the government and educational policy makers. Thus, through data-driven accountability, principals monitor agents' performance and ensure alignments in their aims and incentives to make sure they work to meet their (i.e. principals) set goals.

In this study, we regard the government, Ghana education service, parents and educational policy actors as the core “principals” in the education system, while secondary school teachers are the key agents. That is, in this study secondary-school teachers are the key agents or groups that act (i.e. teach secondary school students) on behalf of Ghana education service, parents, educational policy actors and the government. Because most Ghanaians measure the quality of education by number of students who pass national mandated high-stakes testing, for political reasons it appears successive governments are more concerned about students passing these national mandated large-scale testing at all costs (Baidoo-Anu et al., forthcoming). The desire for students to pass the exams at all costs may bring a misalignment between political aims of the government (i.e. students passing national mandated exams) and teachers who want to achieve the intended learning goals or objectives by promoting formative assessment practices. This misalignment introduces “principal–agent problem”. Thus, misalignment in the aims and incentives (i.e, the need to pass national large-scale exams at all costs) between teachers and governments or educational policy makers results in “principal-agents problem” between teachers and educational policy makers (Jerrim & Sims, 2021). To solve this “principal–agent problem” in education, Figlio and Loeb (2011) stated that government and educational policy makers (i.e. principals) uses data-driven accountability to force teachers (i.e. agents) to align their goals and objectives to the goals and objectives of the government and educational policy makers. Figlio and Loeb (2011) emphatically stated;

The information content in school accountability systems can provide a powerful mechanism for overcoming the principal-agent problem. Assessing schools against the common metric of standardized student test scores provides policy makers and members of the general public with independent information regarding how well schools and school districts (and potentially teachers) are doing in comparison to their peers and outside performance standards. Measuring and reporting school performance and attaching positive and negative consequences to meeting or failing to meet performance objectives provides incentives that encourage educators to concentrate on the subjects and materials that are being measured and to potentially alter the methods through which they educate students. The measurement and reporting of a school’s progress allow policy makers to assess how successful a school has been in meeting the state’s achievement goals (p. 386).

Gleaning from the theory, school-based accountability is premised on the idea that public pressure from openly reported data will lead to an enhanced student success and achievement (Figlio & Loeb, 2011; Rothstein et al., 2008; Supovitz, 2009). However, extant literature explicates that data-driven accountability has unintended negative consequences on the agents (i.e. teachers) (Grupe & Nitschke, 2013; Smith & Amick, 1989). For example, Smith and Amick (1989) found that accountability puts too much stress on teachers. Similarly, Grupe and Nitschke (2013) maintained that close monitoring and punishments attached to accountability leads to stress and anxiety. Further, data-driven accountability has also been found to cause loss in job autonomy (Jerrim & Sims, 2021). Thus, teachers do not have the liberty to work due to data-driven accountability. Finally, data-driven accountability forces teachers to do more work such as preparing students for tests (Perryman & Calvert, 2019). This adds extra work to teachers and puts them under pressure, consequently affecting classroom teaching and learning.

Methods

Research design, sampling and participants

Cross-sectional survey design was employed in this study. Desiring to give participants an equal chance to be part of the study, simple random sampling, specifically the lottery method was used to select the teachers from the three secondary schools in the Birim Central Municipal in Ghana. The secondary schools were also randomly selected using lottery method. In each school, one teacher was trained and engaged as a research assistant to assist with the administration and data collection. During the administration of the instruments, respondents were informed on the purpose of the study and the need to provide honest responses to the items since the study is for academic purposes and also be useful to the school, teachers and students. Teachers were assured that their responses would be kept confidential, and their identity kept anonymous. They were also made to sign the consent form and were apprised that they can withdraw from the study anytime without consequences. It took 7 weeks to collect the data from the respondents. Out of the 227 participants contacted, 200 secondary school teachers completed and returned the questionnaire. This shows 88.1% response rate for this study. Teachers were prompted to respond to survey items based on their individual experience. Out of the total participants, 50 (25%) were females and 150 (75%) were males. Majority 154 (77%) of the teachers had a bachelor’s degree. Moreover, regarding teaching experience, preponderance of the teachers indicated that they have taught for within 6–10 years. This is evident that teachers in this study are experienced teachers who are in the position to respond to the question and provide credible information concerning large-scale accountability influence. The analysis of teachers demographics is presented in Table 1.

Table 1. Demographics of the participants.

Demographic Variables	Frequency	Percentage (%)
Gender
Female	50	25
Male	150	75
Educational Background
Diploma	0	0
Bachelor’s Degree	154	77
Master’s Degree	46	23
PhD Degree	0	0
Years of Experience
0–5 years	20	10.0
6–10 years	166	83.0
Above 10 years	14	7.0

Instruments and data analysis

Teachers’ perception of large-scale testing accountability inventory was developed and used in this study. The inventory was made up of 20 items that sought to understand teachers’ perception of the influence of large-scale testing accountability on teaching and learning. To provide validity evidence, the underpinning conceptual dimensions for the survey were identified based on (a) prior research study, focusing on influence of high-stake testing curriculum implementation in Ghana (Amoako, 2019), and (b) previous surveys and literature on how high-stake testing impacts teachers’ classroom practices (Hamilton & Koretz, 2002; Pedulla et al., 2003; Ritt, 2016; Stecher, 2002) as well as (c) influence of data-driven accountability on teachers’ classroom practices (Figlio & Loeb, 2011; Jerrim & Sims, 2021). Further, three expert researchers were recruited to provide open-ended feedback as well as edits on the wording of items of the questionnaire. The expert further streamlined the items in alignments with research questions and enhanced the clarity of the item wordings. The final survey was pilot tested to check understanding and vagueness and fix any errors that would be due to the framing and development of the items. After the pilot-testing, five items were removed.

In order to understand the structural patterns emanating from teachers’ responses to the influence of large-scale testing accountability on teaching and learning, exploratory factor analysis was conducted. Before starting the factor analysis, data were inspected to ensure suitability for factor analysis. Principal axis factor analysis was employed because it is legitimate for short-scale response options and small sample sizes (Bandalos & Finney, 2018). Oblique rotation (oblimin) was used to examine the component correlation matrix. Exploratory factor analysis was conducted on the 20 items that examine teachers’ perception of large-scale accountability on teaching and learning. The Kaiser–Meyer–Olkin see Table 2) measure validated the sampling adequacy for the analysis, KMO = .634 (“middling” according to Hutcheson & Sofroniou, 1999). This means the sample was adequate for carrying out factor analysis. Bartlett’s test was significant (.000) indicating that the assumption of sphericity was met.

Table 2. Kaiser–Meyer–Olkin measure of sampling adequacy.

KMO and Bartlett’s Test
Kaiser-Meyer-Olkin Measure of Sampling Adequacy.		.634
Bartlett’s Test of Sphericity	Approx. Chi-Square	1001.581
	df	190
	Sig.	.000

Moreover, a preliminary analysis was done to obtain eigenvalues for each factor in the data. Seven factors had eigenvalues over Kaiser’s criterion of 1 explaining 64.04% of the variance. In order to determine the best factor structure pattern, Monte Carlo PCA for Parallel Analysis (Watkins, 2000) and scree plot (see figure 1) were used. The results of this analysis supported three factors, explaining 38.54% of the variance and with eigenvalues more than the corresponding criterion values in a random data set, generated with a matrix of the same size (20 variables × 200 participants).

Figure 1. Scree plot

Factor loadings below 0.4 were suppressed because they did not represent significant values (Field, 2013). Items (12, 13, 7, 5 and 3) did not load onto any factor and were thus deleted. The factors were named based on the items that clustered around the three factors. All the three factors had an internal consistency (i.e. Cronbach’s alpha) greater than 0.70, indicating good reliability (Field, 2013). The three-factor structure was used to understand teachers’ perceived influence on large-scale testing accountability on teaching and learning.

Regarding research question two, participants responses to open-ended questionnaires were thematically analysed. First, the authors organised and (re)read the data. Responses were coded suggestions-by-suggestions, and a code list was generated and refined through an iterative process. Codes were grouped into subthemes in line with our research questions and then into larger themes to describe teachers suggestions. A second ratter was used to review the code list and code 10% of the data. Disparities in the data were co-discussed by the authors until we reached a consensus.

Results

Research question one

Table 3 presents the factor structure and loadings, reliability coefficients (i.e. Cronbach’s alpha), and variance explained. A three-factor structure was identified to understand teachers’ perceived influence of large-scale testing accountability on teaching and learning. Factor 1 which was termed as motivation to teachers centred on six items that imply that teachers perceive large-scale testing accountability as a way of motivating teachers and students. Teachers perceive large-scale accountability as a way of boosting their morale in school. Also, large-scale testing motivates previously unmotivated students to learn. Factor two was termed pressure on teachers. This factor was made up of five items that implies that large-scale accountability puts pressure on teachers. Teachers perceive that large-scale testing accountability forces them to teach to the test. Teachers further indicated that they feel pressure from headteachers to raise students’ scores. Moreover, school administrators and parents also put pressure on them to increase students’ scores. Finally, factor three was termed pressure on students. The third factor was made up of three items. Teachers indicated that large-scale accountability put pressure on students, thereby increasing student test anxiety. Teachers also spend extra time preparing teachers for the test, which consequently increases the number of instructional hours. The reliability estimates for the four factors ranged from 0.702 to 0.74: (a) motivate teachers and students: α = .743, (b) pressure on teachers: α = .702, (c) pressures on students: α = .714. Overall, these three factors accounted for 38.54% of variance in teachers’ perceptions of large-scale testing accountability.

Table 3. Exploratory factor analysis of perceived influences of large-scale testing accountability (high-stake) on teaching and learning.

Items	Factors
	Motivate Teachers and students	Pressure on Teachers	Pressures on Students
14. Large -scale testing has increased teachers’ morale in my school	.745
1. Scores on the large-scale testing perfectly reflect the quality of education students have received	.715
2. The large-scale testing motivates previously unmotivated students to learn	.702
15. High stake test is an appropriate way to assess what student know	.544
16. The national curriculum is aligned with the large – scale testing	.488
17. Ensure teachers are evaluated based on their students’ scores on high-stakes tests.	.446
9. Teachers in my school want to transfer out of form three where the large -scale testing is administered			.716
11. Administrators in my school believe students’ large-scale testing scores reflect the quality of teachers’ instruction			.676
10. Teachers feel pressure from parents to raise scores on the large- scale testing			.661
8. There is so much pressure for high scores on the large- scale testing that teacher only teach to the test.			.554
6. Teachers feel pressure from the headmaster to raise scores on the large – scale testing			.508
18. The use of large-scale tests has resulted in an increase in students’ test anxiety			.697
20. On average, teachers spend at least half of their instructional time (or more) preparing their students for large-scale tests			.647
19. Large-scale testing programme has led teachers to teach in ways that contradict their own ideas of good educational practice			.401
Variance explained % Reliability	18.28 0.743	11.82 0.702	8.44 0.714

Extraction method: Principal Axis Factoring. Rotation method: Oblimin with Kaiser Normalisation. Factor loadings lower than .4 were suppressed.

Research question two

An open-ended question was added to allow teachers to suggest ways that large-scale assessment can be used. The question was “Please, use the space provided below to suggest ways that large scale testing can be used effectively”. Teachers’ responses to this question have been presented in Table 4. Teachers suggested that large-scale testing should not be seen as the sole basis for teacher effectiveness. Rather, they indicated that large-scale testing results should be used to foster equitable distribution of educational resources and foster teachers’ professional development to enhance teaching and learning. Moreover, teachers indicated that large-scale testing results should be used to investigate school and educational policies related to curriculum and instructional practices. Teachers further suggested that large-scale testing results should be reported in a manner that promotes professional collegiality and foster unity among schools, teachers and students.

Table 4. Teachers’ suggestions for effective use of large scale-testing.

Suggestions
Results of large-scale testing should not be seen as the sole basis for teacher effectiveness
Items on the large-scale testing results should be used to foster equitable distribution of educational resources.
Large scale testing results data should be used to improve teaching and learning.
Large scale testing results should be used to examine school policies related to curriculum and instruction.
Large scale testing results should be used to foster teachers’ professional development to enhance teachers’ instructional practices.
Large scale testing results should be used in attracting additional resources for students with greatest needs
Large scale testing results should be reported in a way that promotes professional collegiality

Discussions

Undergirded by theory of the principal–agent problem, the study investigated Brim municipality secondary school teachers’ perceived influence of large-scale testing accountability on teaching and learning. Specifically, the study sought to (a) understand the perceived influences of large-scale testing accountability on senior high-school teachers teaching and learning in Birim central Municipal and (b) teachers suggestions on effective ways to use data from large-scale testing.

Regarding perception of teachers on large-scale testing accountability, the study found that teachers perceive large-scale testing accountability as way to increase their morale in schools and motivate previously unmotivated students. However, they indicated that they feel pressured by the use of data from large-scale testing as a measure of teachers and schools quality and effectiveness. According to the theory of “principal–agent problem” explanation to data-driven accountability, large-scale testing accountability is intended to encourage teachers’ productivity and enhance student success and achievement (Figlio & Loeb, 2011; Rothstein et al., 2008; Supovitz, 2009). Moreover, large-scale testing also offers students with the knowledge of their own progress in their learning and teachers with a deeper understanding of students’ progress and effectiveness of their own instructional practices (Carnoy & Loeb, 2002; Hamilton & Gonzales, 2003; Roderick, Jacob & Bryk, 2002). Notwithstanding large-scale accountability has unintended negative consequences on the agents (i.e. teachers) (Grupe & Nitschke, 2013; Smith & Amick, 1989). Similarly, Grupe and Nitschke (2013) maintained that close monitoring and punishments attached to accountability leads to stress and anxiety. Further, data-driven accountability has also been found to augment feeling of more responsibility, less job autonomy and authority to decide quality practices and teacher effectiveness (Jerrim & Sims, 2021; Prytula et al., 2013). Thus, teachers do not have the liberty to decide which teaching practice work best for their students due to data-driven accountability.

The findings further confirm the work of McMillan, (2005) who found that when large-scale testing results are used for making high stake decisions, it forces teachers to teach to the test, with the primary purpose of increasing students’ scores on external tests. The results further support the work of Kellaghan et al. (2009). According to Kellaghan et al. (2009) large-scale testing puts pressure on teachers to improve student achievement and fear of losing their contracts if they fail to make academic gains. This result further resonates with previous research (Browne, 2016; Berliner, 2011; Kellaghan et al., 2009; Prytula et al., 2013), who found that teachers and headteachers are also under constant pressure to show parents that they are preparing their children to be successful in a test that has the ability to open or otherwise the door to the next level of education.

Teachers in our study also indicated that large-scale testing accountability does not only affect teachers but adversely affect students. The findings support the works of researchers (Ho, 2012) who found that high-stakes test burden and put psychological stress on students, often leading to feeling of failure in life (Baidoo-Anu & Adomaa Acquah, 2021). For example, in China more than 80% of Chinese students recounted worries about high stake testing, leading to headaches and abdominal pain among students (Hesketh et al., 2010). Similarly, high stake testing has been found to increase students’ probability of educational failure, hold students and teachers with unequal and inequitable resources to the same standards, and garbles with the curriculum, and congeal socioeconomic class and ethnic inequalities (Nichols & Berliner, 2008; McMillan, 2005; Burger & Krueger, 2003; Bracey, 2000). Reviewing how large-scale testing shapes students’ motivation, Harlen and Deakin Crick (2003) found that large-scale testing has “particularly strong and devastating impact” (p. 196) on learners who are underperforming. Researchers contend that large-scale testing accountability communicates information that denotes that the main objective of learning is to have high scores on tests (Berliner, 2011; Klenowski, 2011; Klinger & Rogers, 2011; Burger & Krueger, 2003; Bracey, 2000; McMillan, 2005; Nichols & Berliner, 2008). No matter the perceptions others have about the value and utility of large-scale testing, it is indisputable that these testing programmes have had substantial effects on students, teachers, and school administrators. For example, some states developed large-scale assessments that hold head teachers and students accountable. In these systems, students do not graduate unless they pass an exit exam and headteachers do not get raises unless student achievement improves to a predetermined level. Headteachers feel the pressure to improve student achievement and fear losing their contracts if they fail to make academic gains (Berliner, 2011; Kellaghan et al., 2009; Klenowski, 2011; Prytula et al., 2013). Similarly, other researchers point out that when schools are held answerable for student attainment, performing schools can be given recognition and award and failing schools can be identified for additional assistance and resources (Elbousty, 2009; Perkins & Wellman, 2008; Jones & Egley, 2007; Marchant & Paulson, 2005; Stecher, 2002). Suggesting ways of improving the use of large-scale testing, teachers indicated that large-scale testing should not be seen as the sole basis for teacher effectiveness and school quality. Rather, large-scale testing results should be used to enhance equitable distribution of educational resources and foster teachers’ professional development to enhance classroom practices. Teachers are concerned about how the results from large-scale testing are used to evaluate or determine schools and teachers’ effectiveness. Teachers stated that using large-scale testing results to make high-stake decisions create pressure that forces them to teach to test. Investigating the effect of high stakes testing in Ghana basic education, Amoako (2019) reported that teachers engaged students in a narrowed curriculum because of the accountability pressure attached to Basic Education Certificate Examination (BECE). Moreover, Oduro (2015) argued that the majority of teachers conceive assessment as mainly for accountability purposes that tend to obscure the improvement (formative) function. This view of assessment has restricted teachers in exploring ways of involving their pupils in the assessment process and making them take ownership of their own learning.

Implications for policy and practice

This study has accentuated the need for government and educational policy makers in Ghana to rethink the use of data from large-scale testing to make high stake decision regarding teachers and students at the secondary school level. When these large-scale testing are used for making high-stake decisions about teachers and students, it does not only put pressure on teachers but also on students because data from national large-scale testing are also used to make high-stake decisions about students. Volante (2006) maintained that the advantage or disadvantage of large-scale testing lies in its use. Therefore, to leverage large-scale testing, results should be used to examine school policies related to curriculum and instruction and promote better allocation of state educational resources. The fundamental value of large-scale testing programmes is contingent on their capability to promote, and not hinder student and teacher learning. To do this, large-scale testing must be seen as significant, but not as measures for high-stakes decisions about teachers, students and schools (Volante, 2006). Therefore, Ghana Education Service should consider making large-scale testing low stakes. Data from large-scale testing could be used to foster better allocation of educational resources and encourage teachers’ professional development to improve instruction. Again, evaluation of school and teachers’ performance on the basis of students’ performance on large-scale testing creates tension and puts pressure on teachers to teach to the test that eventually narrows the curriculum (Amoako, 2019). Hence, to avoid unhealthy competition among schools and teachers, LSA results should be reported in a manner that encourages professional collegiality. Educational policy makers and politicians should not assume that large-scale testing alone can on its own increase teachers’ productivity and students’ achievement (Black & Wiliam, 1998; Kohn, 2000). Rather, encouraging and supporting teachers to engage in quality classroom assessment has the potential of moving learning forward and increasing students’ success. Merely providing teaching and learning materials (Adane, 2013; Baidoo-Anu, 2018a, 2018b), educational infrastructures (Baidoo-Anu & Mensah, 2018), would have little impact on students’ overall academic performance if teachers do not engage in assessment practices that move learning forward (Black & Williams, 2010). Irrespective of the assessment literacy of teachers, if large-scale testing results are not used appropriately, it will negatively affect teachers’ assessment practices (Volante, 2006). To avoid wash back (i.e. teaching to the test) (Carless, 2011), and other negative effect of large-scale testing, educational stakeholders in Ghana should eschew using large-scale testing results for making high stake decisions like solely determining teachers’ and schools’ effectiveness, rewarding or punishing teachers and schools, streaming or tracking students, determining students educational progress etc.

Limitation and suggestion for future research

The sampling of a small size in one municipality in Ghana, however, limits the generalisability of the study’s findings, which is a limitation of the study. Again, the study relied largely on questionnaires as the main data collection instruments can introduce biases. This can happen for a variety of reasons, including social desirability bias and attempting to protect privacy. However, assuring participants of anonymity and confidentiality of their responses helps reduce the probability of this problem occurring. Notwithstanding, this study has provided some unique compelling evidence. Future research could consider increasing the sample size to allow for broader generalisation of the results.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

We have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Notes on contributors

David Baidoo-Anu

David Baidoo-Anu is currently a PhD in Education Candidate at the Faculty of Education, Queen’s University in Kingston, Ontario Canada. He earned his Master of Philosophy in educational measurement and evaluation from the University of Cape Coast-Ghana. Previously a part-time lecturer at Presbyterian University College- Ghana, his enormous interest in research, particularly classroom assessment coupled with his experience directed him to the Queen’s Assessment and Evaluation Group (AEG). He believes that if classroom assessment is appropriately used, it has the power to support equitable access to education and propel learning forward for diverse students. David’s current research focuses on culturally responsive classroom assessment. Particularly, he draws on culturally situated understandings of assessment and empirical evidence showing the positive impacts of quality classroom assessment practices on raising students’ achievement to advocate for systemic assessment reforms. The aim of his doctoral research is to support quality and consistent assessment practices in schools across cultural context and ultimately enhance the assessment experiences of all students regardless of their background. Isaac Ennu BaidooIsaac Ennu Baidoo is a mathematics teacher at Atweaman Senior High School in Akyem Manso-Ghana. He is also a senior Tutor at the University of Cape Coast Distance Education Program. He holds Master of Education in Teacher from University of Cape Coast- Ghana. Isaac is currently completing Master of Philosophy in educational measurement and evaluation from the University of Cape Coast-Ghana. Isaac’s research focus on the influence high-stake testing on teaching and learning with specific interest in mathematics.

Isaac Ennu Baidoo

Isaac Ennu Baidoo is a mathematics teacher at Atweaman Senior High School in Akyem Manso-Ghana. He is also a senior Tutor at the University of Cape Coast Distance Education Program. He holds Master of Education in Teacher from University of Cape Coast- Ghana. Isaac is currently completing Master of Philosophy in educational measurement and evaluation from the University of Cape Coast-Ghana. Isaac’s research focus on the influence high-stake testing on teaching and learning with specific interest in mathematics.