The Impact of Strategy Type on the Relationships among Broad Scope Management Accounting System, Job Information Perceptions, and Managerial Performance

Chiao Da }vfanagement Revi阱， 均1. 29 No. 2, 2009

pp.47-82

策略型態對廣範園管會系統、工作資

訊認知與管理績效之關聯性的影響

The Impact of Strategy Type on the Relationships among

Broad Scope Management Accounting System

,

Job

Information Perceptions

,

and Managerial Performance

鋒紹熙 1 Shao-Hsi Chung

美和技術學院 企業管理學系

Departmenl ofBusiness Administralion, Mei-Ho Instilule ofTechnology 倪豐裕 Feng-Yu Ni

國立中山大學 企業管理學系

Dep叮tmenlofBus剖tn】ess Admini昀Sl叮ra剖\1叩on ， Na瓜11叩on訓1a Sun Ya刮I-Se叩n Unive叮rs剖1t句y 豆緣車英芳 Y泊1呵r

圓立中山大學 企業管王理重學 j象

_h

PhD. ofBusiness Adminis甘alion ， Nalional Sun Yal-Sen University 邱炳乾 Bing-Chyan Chiou

國立屏束商業技術學院 財務金融系

Departmenl of Finance, Nalional Pingtung lnslitule of Commerce 蘇錦俊 Chin-Chun Su

高苑科技大學 國際商務系

Departmenl oflnlemalional Business, Kao-Yuan University

摘要: 本研究主要目的 ， 在探討廣範園管理會計系統與管理績效之闖餘，是

否受到工作攸關資訊認知的中介影響，以及廣範函管理會計系統的使用程

度、廣範園管理會計系統和工作攸關資訊認知以及工作攸關資訊吉思知和管理

I Corresponding author: Departrr>ent of Business Adminis甘ation， Mei-Ho Institute of TechnoJogy, Pingtung, Ci吟，而WaJ1. E-m但只3J 66

@m

eiho.edu.tw

48 The lmpacl ofSlralegy砂'Peon the Relalionships among Broad Scope Managemen1 Accounting Sysfem, Job Informaliol1 Perc申I;OI1S， and Managerial Performance

績效的聯結關餘，是否受到不同策略型態的影響 。 為驗室查這些關餘 ， 本研究 使用路徑分析法，並採問卷調查法進行研究，研究對象伶、台灣地區電子產業 上市(櫃)公司的廠長及其下屬員工 。 研究結果發現，廣範園管理會計系統 對於管理績效之影響，並非直接而你完全經由工作攸關資訊認知的中介效 果，尤其在企業傾向差異化策略時 ， 此完全中介效果會更為明顯 。 關鍵詞:廣範園管理會計系統;工作攸關 資訊認知;管理績效;策略型態

Abstract: The main purpose ofthis study is to examine the effect ofperception of job- relevant infonnation (PJRI) 的個 intervening variable between broad scope

management accounting system (broad scope MAS) and managerial perfonnance Further, this study also examines the impact of strategy type on the level of using broad scope MAS and the relationships between broad scope MAS and PJ悶， and between PJRI and managerial perfonnance. The results, based on path analysis and a questionnaire survey of plant managers and their subordinate in Taiwan's public electronic fi口ns， demons甘ate that broad scope MAS a釘ècts managerial perfonnance, not directly, but completely through PJRI. In particular, the full mediating e前的t of P JRI is more pronounced when finns tend to adopt a differentiation strategy. However, when they tend to adopt the low-cost strategy,

the relationship bet、lVeen broad scope MAS and P成1 disappears

Keywords: Broad Scope Management Accounting System; Perception of

Job-Relevant Infonnation; Managerial Perfonnance; Strategy Type

1.

Introduction

In face of technological advancement and global competitions, corporate managers are working in an increasingly complicated environment. To ∞pe with this highly competitive environment, they should be given adequate infonnation (Galbraith, 1973). Management accounting system (MAS) is intended to help managers access and use management accounting infonnation necessary for accomplishing goals of their work, so it can e仔èctively enhance their managerial perfonnance (Baines & Langfield-Smi曲， 2003; Chong & Chong, 1997). However,

Chiao Da Managemenl Revi.飢似的1. 29 No. 2, 2009 49

the simple and direct relationship between MAS and managerial performance has been questioned by many previous researchers (Chenhall & Brownell, 1988; Gul, 1991; Mia & Chenhall, 1994). They have provided two different perspectives. In one perspective, the relationship between the two is considered as moderated by certain contextual factors, such as environmental unce此ainty， task uncertain旬， organizational struc仙時， and job properties (Chi

a,

1995; Chon皂， 1996; Gul & Chi

a,

1994; Mia & Che叫】a1 1， 1994). In the other, it is influenced by certain intervening factors, such as perception of role ciarity, perception of job-relevant information, and perception of job stress (Hall, 2008; Kren, 1992; Shields, Deng

& Kato, 2000)

An MAS provides information necessary to decision making, inciuding ∞st ac∞unting information, product pric巴， and profit information (Tiessen &

Waterhouse, 1983). The information it provides can be ciassified into broad scope information and narrow scope information. The broad scope MAS info叮nal10n inciudes extemal (e.g. customer preference, competitor activi秒， and technology development), non-financial (e.g. yield, defect rate, and machine operating efficiency) ，個dfu仙re-oriented information (e.g. probability that a new law is legislated), while the narrow scope MAS information inciudes intemal, financial,

個d historical information (Chenhall & Morris, 1986). This study focuses on the broad scope 恥的 S information and uses perception of job-relevant information (PJRI) as a mediator variable to explore the relationship between broad scope MAS and managerial performance. Besides, the moderating e能ct ofs甘ategy type (differentiation and low-∞st) on the relationship between broad scope MAS and managerial performance is a1so examined. Thus, the motivation behind this study is to investigate the practical ∞rrelation between broad scope MAS and managerial performance using a mediator variable (PJRI) and a moderator variable (strategy type). The objectives ofthis study are as follows: Theoretically,

we aim to construct a relationship model of broad scope MAS and managerial performance with consideration of a mediator variable and a moderator variable. Practically, we expect to provide enterprises with some suggestions on how to make use ofbroad scope MAS to improve managerial performance

50 The 1m月clojSlralegy 砂'pe on the Relalionships among Broad Scope Management Accounling Syslem, Job 1njomwlion Perceplio肘， and Managerial Perfonnance

hypotheses are presented in Section 2. Research methods are explained in Section

3. Empirical resu1ts and analysis are detailed in Section 4. Finally, conclusions

and research limitations are proposed in Section 5.

2.

Literature Review and Hypotheses

2.1 The Relationship Between Broad Scope MAS and

Managerial Performance

To find out whether broad scope MAS can effectively e的 ance managerial performance, previous researchers usually adopted the contingency theory to test the e釘ectiveness of broad 5∞pe 恥1AS， with no consideration of the effect of any mediator. For instar肘， Mia and Chenhall (1994) examined 出 e managerial performance of different functional dep訂的lents of an enterprise using broad

scope MAS. They argued that broad scope ~也屯S is more effective for the marketing department 由an for the production depar的lent. Their argument was

empirically supported. Gul and Chia (1994) explored the effect of perceived environmental uncertainty, decen甘alization, and broad scope MAS on managerial

pe巾rmance. Results of their 5ωdy indicated that only when the perceived

environmental 叫“巾inty is hi阱， use of decentralization and broad 5∞pe MAS has a positive effect on managerial performance. Chong (1996) investigated the

effect of broad scope MAS on managerial performance at di叮erent levels of task uncertainty. Results revealed that under high task uncertainty situations, high extent of use of broad s∞pe MAS could lead to improved managerial performance; under low task uncertainty situations, high extent of use of broad

s∞pe MAS would cause information overload and reduction of managerial

performance. 肉， Chuang and Chiou (2006) studied the effect of broad scope MAS on managerial performance with consideration of organizational commitment and task unce巾inty. They found that heavy use of broad scope

MAS could enhance managerial performance only when organizational

Chiao Da Managemenl Review Vol. 29 No. 2, 2009 51

2.2 The Mediation ofPJRI

Many researchers of management accounting have pointed out that the

effects ofmany MAS-related antecedent variables (e.g. performance measurement

system, budgetary participation) on managerial performance require mediation of

psychological perceptions (e,g. budgetary goal acceptance, perception of role

cI ar旬， perception of job-relevant information, and perception of job 泊的s) (Hall, 2008; Kren, 1992; Shields, Deng & Kato, 2000; Zhu, Lin & Ni, 2003). For instance, Hall (2008) confirmed that the relationship between performance measurement systems and managerial performance is mediated by perception of role cIarity. Kren (1992) argued that budgetary participation influences managerial performance through PJR1 and also empirically validated this argument. However, so far, broad scope MAS has never been used as an antecedent to discuss how its relationship with managerial performance is affected by mediator variables. Managers usually apply broad scope MAS information to task planning and decision making (Chenhall, 2003). In 出is study, we define information provided by broad scope MAS as job-relevant information 2 to

examine the relationship between broad scope MAS and managerial performance Besides, as suggested by Kren (1992), P JR1 is a key factor affecting a manager' s

use of MAS to enhance hislher performance. Hence, we set PJR1 as a mediator

variable to examine its relationships with the antecedent variable (broad scope

MAS) and the outcome variable (managerial performance)

PJR1 refers to managers' perception of the degree to which they can receive task-relevant information (Kren, 1992), such as whether they understand how to

ac∞mp1ish a certain tas

k,

whether they have sufficient information to make an optimal decision, and whether they can obtain important information to assist

their decision of an impo此ant solution. Therefore, PJR1 is a belief of managers

2 ln previous s仙刮目。f management ac∞ unUn且. org缸U胡tional information was considered 自

inclusive of decîsion-facilitating information and decision-inf1uencing infonnation

Decision-facilitating infonnation is the infonnation ∞llected prior to making a decision Decision-inf1uencing information is related to manager祖I behavior. lt is mainly used to evaluate

a manager's behavioral performance (8副m凹， 1982; Tiessen & Waterhouse, 1983). Kren (1992) defined decision-facilitating information as job-relev阻t information. Williams and Seaman (2002) generalized.the two kinds of information as managerial relev酬 information

52 The 1m戶ctojStrategy 砂pe on the ReJatìonships among 8road Scope Management Accounting System, Job lnjormation Perceptions, and Managerial Performance

about whether they can obtain infonnation necessary for evaluating, pl街mm皂， and

executing task decisions to accomplish the goal oftheir jobs

From a broad scope MAS, managers can retrieve broad scope informatio

n,

including fuωre events (e.g. new legislations), external situations (e.g. threat from competitors, economic conditions), and non-financial infonnation (e.g. market share and yield). Such infonnation can make them believe that 由ey have obtained sufficient infonnation for planning and executing task decisions (Chon皂， 1996). In other words, broad scope MAS infonnation elevates the degree to which managers can obtain job-relevant infonnation ac∞rding to their per∞ptions Therefore, we argue that there is a positive relationship between broad scope MAS and PJRI

Provided wi由 sufficient infonnation related to the environment and j鉤，

managers can reduce the interference of infonnation gap and role ambiguity to

E的個ce their job perfonnance (Che他叫 1 & Brownell, 1988; G剖 braith，

1977). PJRI and role ambiguity are based on a similar concept (Kr凹， 1992). It

C個 be inferred that 的RI also has positive and substantial inf1uence on managerial

perfonnance. Thus, we propose that there is a positive relationship between P JRI and managerial perfonnance. In sum, there is no direct relationship between broad scope MAS and managerial perfonnance, and P民J plays the impo此ant role as a mediator in 曲的 relationship. Jt is a key factor that detennines whether use of broad scope MAS can enhance managerial perfonnance. As illustrated in Figure 1, the following hypotheses are proposed.

H 1: The relationship between broad scope MAS and managerial performance is

mediated by P JRl.

H1Q: There is a positive relationship between broad scope MAS and P JRI

H1b: There is a positive relationship between PJRI and managerial perfonnance (when both broad scope MAS and PJRI are independent variables of managerial performance)

H1C: There is no direct relationship between broad s∞pe MAS and managerial

perfonn個ce (when both broad s∞pe MAS 個d P JRI are independent variables of managerial perfonnance)

Chiao Da Management Review Vo/. 29 No. 2, 2009 53

Figure 1

The Relationships among Broad Scope MAS

,

PJRI

,

and Managerial Performance

PJRJ

8road Scope MAS M缸lagenal Perfo口n缸lce

H

2.3 The Influence of Strategy Type on Relationships among Focus Variables

Strategy refers to the courses of action and the allocation of resources necessary for carrying out the long-term goa1s of an enterprise (Chandler,

1962). Previous researchers have proposed varying typologies of business

strategies. For instance, Mi1es and Snow (1978) c1assified strategies into prospector strategy, defender strategy, and analyzer strategy, according to market characteristics. Porter (1980) proposed three competitive strategies, name1y

di釘erentiation， overall cost leadership, and focus. According to Govindarajan and

Gupta (1985), corporate strategies should be positioned on a continuum bet、veen

build and harvest. In academic research and corporate practices, PO此前's typology (1980) is most commonly adopted. For example, Govindarajan and Fisher (1990)

applied di叮erentiation and 10st-∞st strategies to explore the effects of strategy, control systems, and resource sharing on business unit perfo口nance. Van der

Stede (2000) a1so used the same strategies to examine the effects of strategy type

and past performance on budgetary control, budgetary slack, and manageria1

short-term orientation. Accordin駒， we also c1assify co巾。rate strategies into differentiation strategy and low-cost strategy in 出 is study. Due to 出e fact that management accounting information is helpful for selection and execution of strategl的， there is a c10se relationship between use of MAS and strategy type (Chenha11, 2003). To understand the e仟ect of strategy type on the level of using broad scope MAS and relationships between broad scope MAS and PJ間， and between P JRl and managerial performance, we will use these two strategies to

54 The lmpacl of Slralegy 砂peon the Relationships among Broad Scope Management

Accounling 砂'Slem， Job lnfonnalion Perceplions, and Managerial Perfonnance

examine the effect of strategy type on the above-mentioned relationships

In an environment of high uncertainty, enterprises tend to adopt a differentiation strategy in its production (Miller, 1988; Porter, 1985). With an aim to produce unique products, managers need to understand customer preferenc唔， threat from competito時， technology developmen

t,

and other extemal,

non-financial, future-oriented information. For them, broad scope management accounting information is urgently needed (Baines & Langfield-Smith, 2003). In the study of the effect of environmental un∞巾inty on budgetary participation and the effect of budgetary participation on managerial performan白， Kren (1992) discovered that under environmental uncertain旬， manage時， high degree of budgetary participation can increase 間間， which further enhances managerial performance. Besides, Williams and Seaman (2002) probed into the effects of

management accounting and control system changes on departmental

performan∞. They validated the full mediating effect of perception of managerial relevant information and found that it becomes more pronounced ' when job uncertainty is high. Therefore, we argue that when ente叩rises tend to adopt a differentiation strategy, their managers will have a high level of use of broad scope MAS. Besides, despite the adoption of the differentiation strategy, there is no direct relationship between broad scope MAS and managerial perform翻白

When a differentiation s恤tegy is employed, managers' heavy use of broad scope MAS can help them acquire necessary broad scope management accounting

information (Bouwens & Abeme血y， 2000). W抽 sufficient broad scope

information, such as 血ose irivolved in budgetary participation, managers will have a higher degree of P 1即如 d thus e油 ibit higher managerial performan∞

(Kren, 1992). It can be infe叮ed that when an enterprise tend to employ a differentiation strategy, broad scope MAS is positively correlated with P1RI,

and PlRI is also positively ∞汀elated with m個agerial performance.

In a relatively more stable environme肘， ente叩rises may tend to adopt the low-cost s甘ategy (Miller, 1988; Porter, 1985). In this stable environmen

t,

production processes are highly standardized and specifi叫， and managers have clear understanding of the properties of their jobs. Therefore, they do not need to highly rely on broad scope MAS (Chong & Chon皂， 1997) to acquire more

Chiao Da Management Review Vol. 29 No. 2, 2009 55

job-relevant information to e伯伯 ce managerial performance (Chong, 1996) 。

Meanwhil巴， high level of use of broad scope MAS would rather cause information

overload, resulting in the loss of focus and increase of load on managers, PlRI

and managerial performance will not be positively influenced. 8ecause most of

the processes have been standardized and speci日叫， tasks can be operated like routines. Managers can only follow existing rules and processes to effectively

accomplish goals of their jobs. Hence, a high degree PJRI has Iimited effect on

improvement of managerial performance (Leblebici & Salanci

k,

1981). It can be inferred that when an enterprise tends to employ a low-cost strategy, the level of

using broad scope MAS will not be high. 8road scope MAS has no effect on managerial performance or P J悶， and P JRI and managerial performance are also not positively correlated to a high extent

In sum, we argue that high level of using broad scope MAS, and the

relationship between broad scope MAS and managerial performance with

consideration of PJRI as a mediator is more applicable to ente巾的的 inclined to employ the di缸èrentiation strategy. In other words, the level ofusing broad s∞pe MAS, the correlation between broad scope MAS and PJ悶， and the correlation

b巴tween PJRI and managerial performance are lower when the low-cost strategy

is employed. As illustrated in Figure 2, the following hypotheses are established:

Figure 2

The Effect ofStrategy Type on Broad Scope MAS and Relationships among

Broad Scope MAS

,

PJRI

,

and Managerial Performance

Strategy Type

H2a

/

(DifferentiationILow-cost)

\

H2b H2c

Broad Scope P1RI Managerial

H2

.

_MAS

low-cost

5 ~rprises ten Perforrnance

帥..日 _•••

te

J!J!:

56 The 1m戶口 ofSlralegy 砂'Peon Ihe Relationshìps among Broad Scope Monogement

Accounting 是ySlem， Job Informalion Perceplio肘， and Managerial Perfomwnce H2b: the positive relationship between broad 5∞pe MAS and PJRI is high釘，

H2C: the positive relationship between PJRI and managerial performance is higher

3.

Research Methods

3.1 Sample Source and Data Collection

The research sarnple comprised ofpublicly listed electronic firms in Taiwan By consulting the prospectus of the selected electronic fi口ns， we obtained the plant address of each firm. Two questionnaires were mailed to only one plant of each firm. Before maili月 the questionnai悶， the researchers telephoned the personnel department of each firm to acquire consent and narnes of the plant manage? and one selected subordinate of the manager. One questionnaire was mailed to 血e plant manager. This questionnaire was designed on the basis ofthree

variables, including broad 5∞pe MAS, PJRI, and strategy type. The other

questionnaire was mailed to the subordinate, who was required to evaluate 由e

plant manager's managerial performance4 using perception measuremenr. The

questionnaires were responded on anonymity and directly mailed to the researcher.

A total of 680 firms were selected at random. Of the questionnaires retumed, 124

were collected from plant managers, and 153 企om subordinates. However, only 107 effective pairs of questionn凹的 were obtained, making the valid response rate 16%. The age, senior旬， and education background of the respondents are 3 In也;spaper, plant manager was selected as a repre甜n個tive ofmanagers of 叩開te叩n甜

4 Among the studies using per臼ived performance 自由e basis of managerial performance,

Brownell and Mclnnes (1986) sugg叫“出at self-evalua包on should be avoided. Besides, J(jm

and y，也1 (1995) mentioned 也at when evalua恤g 出e managerial performance of the superior 由rough perception measurement, evaluation by subordinates wouid be better 曲曲 目lιevaluation of 出e superior. A1though letting a plant manag缸's performance evaluated by hislher superior is 刮目 a reasonable and etfcctive approach in perception measurement. the

superior of the plant manager (i.e. general m個 ager) may be too busy to answer 0叮 questlOnnarre η隨時fo間， considering time and e節目ency， we let the pl曲tm阻ager's

managerial performance evaluated by onJy their subordinates

,

Perception measurement is a way to evaluate the weight of a question according to per回n叫

Chiao Da Manag.ωnent Review Vo/. 29 No. 2. 2009 57

provided in Appendix 1.

3.2 Measures

AII the measures used in this study were extracted from questionnaires developed by previous researchers (see Appendix 2), so the content validity of these measures could be ensured. Besides, reliability of broad scope 恥也屯S

and PJRI was tested using Cronbach u (1951), and the convergent validity and uni-dimensionality of these variables were evaluated using confirmatory factory analysis. AII these variables (excluding strategy type) were designed to be evaluated on seven-point Likert scale. These m巴asures are respectively explained as follows

3.2.1 8road Scope MAS

To measure the extent to which each plant manager uses the broad scope MAS, the scale comprising six items developed by Chenhall and Morris (1986) was adopted (ltem 1 and 2 were focused on future-oriented information, Item 3 and 4 on external information, and Item 5 and 6 on non-financial information). Chenhall and Morris (1986) have verified the reliability and validity ofthe scale. Most of the foreign researchers on broad scope 恥1AS also adopted this sc此

(Bouwens & Abernethy, 2000; Chi

a,

1995; Chong, 1998; Chong & Chong, 1997;

Fisher, 1996; Gul, 1991; Gul & Chi

a,

1994; Mia & Che叫lall， 1994). Besides, the majority of domestic researchers 00 this subject also adopted this scale (Chang &

Chang, 2001; Chang, Chang & Ni 2000; Ni, Chuang & Chiou, 2006; Ni & Su,

2001). In this study, the Cronbach u of this variable was 0.77. In Chenhall and Morris (1986), it was 0.76

3.2.2 Perception of Job-Relevant Information (PJRI)

In this s仙dy， PJRI was measured using the scale developed by Kren (1992)

。n the basis ofinformation overload index introduced by O'Reilly (1980). This scale consisted ofthree items which can be used to evaluate whether the plant manager clearly understands how to accomplish the tasks, whether he has

sufficient information to make optimal decisions to achieve the performance goal,

58 The Impact ofStrategy 砂'Peon the Relationships among Broad Scope Management Accounting System, Job Information Perceptions, and Managerlal Performance

study, the Cronbach a ofthis variable was 0.83. In Kren (1992), it was 0.72. 3.2.3 Managerial Performance

Managerial performance was measured using the 凹的 developed by

Mahon句， Jerdee & Carroll (1965), and the result of the subordinate's perception

measurement was used as a measure of this variable (Abernethy & Brownell,

1997; Chalos & Poon, 2000; 扎針a & Chenhal1, 1994). This scale consisted of eight

sub-dimensions, including planning, investigating, coordinating, evaluating,

supervising, staffing, negotiating and representing as well as one dimension of

overall performance rating. Mahoney et al. (1965) pointed out that the eight sub-dimensions should be independent and account for at least 55% variance of the overall performance rating, wi血 the remammg 45% being explained by

specific job factors of managers. To veri

fY

the independency of these

sub-dimensions, we used the experiential approach suggested by Pindyck and Rubinfeld (1976) to test the multicollinearity of the independent variables (i.e

eight sub-dimensions). This approach defines that multicollinearity occurs if the correlation coefficient between two independent variables is greater 由an the coefficient be仰的n the two variables and the dependent variable (i.e. overall performance rating). In this study, among the 28 pairs, only 4 satisfied this criterion. Brownell and McInnes (1986) and Kren (1992) obtained 3 and 4 pairs respectively, using the same approach. Hence, the eight sub-dimer>sions featured a

certain level of independence

We later conducted multiple regression analysis of the eight sub-dimensions to test their explanatory power for the overall performance rating Result showed R2 = 63%, which exceeded the 55% standard6 R2 was 78% in

Brownell and Mclnnes (1986), 68% in Chia (1996), 60% in Chong (1995), and 57% in Chong (1998). R2 = 63% indicated that the eight sub-dimensions accounted for 63% variance of overall performance rati嗯， and the remaining 37%, as suggested by Mahoney et 叫 (1965)， ∞uld be explained by specific

job-relevant factors of managers. The dimension of overall performance rating

,

Ac∞rding to Brownell 血d McInnes (1986)，出e mam p呻0時 of this test is 10 e曲扭扭曲e

Chiao Da Management Review Vol. 29 No. 2, 2009 59

consisted of 63% explained by the eight sub-dimensions and 37% explained by

specific job-relevant factors. This indicated that the eight sub-dimensions could

be covered if only the overaJl performance rating was measured. Among the

studies focused on the effects of MAS on managerial performance, most of those using the approach proposed by Mahoney et 叫(1 965) employed only one item

(i.e. the overall performance rating) to measure managerial performance

(Abemethy & Brownell, 1997; Brownell & Mclnn的， 1986; Chon皂， 1996; Chong

& Edmund, 2002; Gul & Chia, 1994; Zhu, Lin & Ni, 2003). Therefore, in this

S仙旬， we also measured managerial performance only using overall performance ratmg

3.2.4 Strategy Type

To measure the strategy type that enterprises tend to employ, we adopted the one item approach developed by Govindarajan and Fisher (1990). Every plant manager would measure the percentage of total tumover (in 1 ∞%) gained from differentiation strategy and the percentage from low-cost strategy. We multiplied the percentage from di旺erentiation strategy by 1 and the percentage from

low-∞割的ategy by

.

1. The sum of the two vaJues was then used to denote

strategy type. This value should range between -1 and 1. The c10ser it was to 1,

the more that the enterprise tended to employ the di缸erentIatlOn s甘ategy; the

c10ser it was to -1, the more that it tended to employ the low-cost strategy. We used the median (= 0) to c1assify the samples by strategy type. The numbers ofthe firms c1assified into the differentiation and low-cost strategy groups were 54 and

53. The difference between the two groups was further tested using independent

sample t-test. A significant difference was detected (t = -11.471, P < 0.000).

3.2.5 Evaluation of Uni-Dimensionality

To ensure the uni-dimensionality of broad scope MAS and PJRI, we

conducted confirmatory factor analysis of 出e two variables. A measurement model was constructed and LlSREL was used as an aJytic剖 tool. Results showed 伽t the ove耐 model fit was good

(:l

= 43 鈣， d.f. = 24, p = 0.008，的 d.f. = 1.83,

GFI = 0.92, CFI = 0.94, NNFI= 0.91, RMSEA= 0.088). As shown in Table 1, aJl the standardized factor loadings of the construct variables on items reached the

60 The Impact of Strategy 砂pe on the Relotionships among Broad Scope Management Accounting System, Job Information Pe阿eptions， and Managerial Pe彷nnance

level of significance in the t

.

test (p < 0.01), so the scales of the variables had convergent validity (Anderson & Gerbing, 1988). Besides, a11 the composite reliability va1ues and Cronbach U of the two variables 剖 so exceeded 出e 0.7

criterion, indicating the two variables had an acceptable level of intema1 consistency. Through the above analyses, we verified the uni-dimensiona1ity of broad scope MAS and PJRI.

Table 1

Evaluation of Uni-Mimensionality

S旭n也rdi旦dp缸缸neters

Variable Factor loading Measurement Composite Cronbach tl

(t-value) error reliability

Broad s∞peMAS _{。 74 。 77} BS恥IAS 1 _{。“(一句 。 79} BSMAS 2 0.62 (5.08") 0.62 BSMAS 3 0.61 (3.43'') _0.62 BSMAS4 。 51 (3.11-) _{。 74} BSMAS 5 。 56(3.34 ") 0.68 BSMAS6 。 61 (3.44") 0.63 PJRI 0.85 0.83 PJRl 1 ₀.53 (一') 。 72 PJRl 2 。 96(5.82'') 0.07 PJRl 3 。 90(5.96'') _{。 19}

Note: 'fixed parameter " _{p < 0.01}

3,3 Analysis Method

Path analysis was conducted to explore the causal relationships among broad s∞pe MAS, Pl悶， and manageria1 performance (as shown in Figure 1). If the causa1 relationship between research variables is theoretica1ly supported, verification of 出e relationship through path analysis is considered as appropriate

Chiao Da Managemenl Review Vol. 29 No. 2, 2009 61

The purpose of path analysis is to verify whether the causal model proposed by the researcher fits practical data (Banker, Bardhan & Chen, 2008; OeGroot &

Brownlee, 2006). Before estimating path coefficien郎， data should be standardized. The values are later computed through regression analysis (Asher, 1983). Through

path analysis, correlation coefficient between two variables (total effect) can be decomposed into direct effect and indirect effect or spurious effect to judge if there is any mediator variable between the two variables

4.

.

Empirical Results and Analysis

Table 2 and Table 3 respectively show the descriptive statistics and the

correlation matrix of the variables. As shown in Table 3, there are positive and

slgm日cant relationships between broad scope MAS and 凹悶， and between broad

scope MAS and managerial perforrnance. Also, there is a positive and significant

relationship between P lRl and managerial perforrnance. These findings are

consistent with conclusions in several previous studies (Baines

& Langfield-Smith, 2003; Chong & Chon皂， 1997; Kren, 1992; Williams &

Seaman, 2002).

Table 2

Descriptive Statistics (n = 107)

Measurement variable Mean S.D Theoretic score ActuaJ s∞悶

Broad scupe MAS 30.27 5.38 6-42 17-42

PJRI 15.83 3.11 3-21 7- 21

ManageriaJ 5.43 _{。 79} 1- 7 3-7

perfonnance

62 The Impact ofStrategy 砂'pe on the Relationships among Broad Scope Management

Accounting System. Job Information Pe間ptions. and Managerial Performance

Table

3

Correlatioo Matrix (0 = 107)b

Measurement variablel

Broad s∞ peMAS 。 310"

PJRI Managerial perfonnance

PJRI

Managerial perfonn閻明 。 241 ﹒ 。 475""

Strategy type 0.068 _{。 188}

Note: • all variables have been standardi血d. btwo-恤led test was adopted

""p < 0.01

.

p< 0.05

4.1

The Meditating Effect ofPJRI

。 132

We applied path analysis to verify the effect of broad s∞pe MAS 00 managerial performan∞出rough PJRl. In Figure

3

, ZI denotes broad s∞pe MAS,

已 denotes PJ間， and Z3 managerial performance. AlI these variables have been

standardized. Path coefficients are denoted by Pij (i.e 恥，肘， and P32).

R.

and Rv

represent the unexplained parts in P JRI and managerial performance, respectively

The structural equations are shown in Equation (1) and Equation (2) 。

Z2 =P21 ZI +凡

Z3 =P31 ZI

+

P32 Z2

+

Rv

(1) (2)

The solution of the path coefficients Pij is derived 企om regresslOn an剖 ySlS

and estimated using the ordin訂y least squares method. In fact, P21 is the

standardized regression ∞efficient， while P31 and P32 缸e partial regression

coefficier>ts. Each Pij denotes the inf1uence ofvariable j on the variance ofvariable

i. The variance is expressed by standard deviati凹 的 the unit. When variable j

varies by one unit of standard deviation, variable i will vary by p unit of standard

deviation. Besides, the residual terms Ru and Rv in the equations are hypothesized

as non-correlated with the explanatory variables

The results of Equation (1) and Equation (2) with all samples after

Chiao Da Managemel1t Re叫ew Vol. 29 No. 2, 2009 63

shown in Figure 3. The results indicate that Equation (1) and Equation (2) with all

samples have reached the significant level. F-values are respectively 11.129 (p <

0.01) and 16.033 (p < 0.01). Besides, the path coefficient between broad scope

MAS and PJRl (即 1)， and the path ∞efficient between P JRl and managerial

performance (P32) are significant. However, the direct path between broad scope MAS and managerial performance (P31) is insignificant. Therefore, _H1a, Hlb, and

H1c are supported. This manifests that PJRl plays the role of a mediator in the

relationship between broad scope MAS and managerial performance

Li叫咱 _Path

coefficient

Table 4

Path Analysis Result

AlI samples (n= 107) Differenriation strategy (n = 54) Low-cost strategy (n = 53)

estimate t-value es1Îmate t-value estimate t-value

Equation (1) PJRIIBSMAS Equation (2) Managerial perfonnancelP JR I Managerial perfonnancelBS MAS Equat Equation (2) p" 。 310 3.336** 。 474 3.877" p" 0.443 4.918** 0.477 3.580*' p" 。 103 1.146 。 124 AlI s缸nples， R' = 0.096, F = 11.129** Dìfferentiation strategy. R2 話。 224， F = 15.031" Low-cost strategy, R' = 0.034 ' F = 1.792 AlI s祖nples， R' = 0.236, F = 16.033** Differentiation strategy, R' = 0.298, F = 10.850'* Low-cost strategy, R' = 0.184 ' F = 5.625** 0.930

Note: BSMAS = broad scope MAS. PJRl = perception ofjob-叫evant info盯natIOn

**p < 0.01

。 184 1.339

。 415 3.191*

*

64 111e lmpact ofStrategy 砂'pe011 the Relationships among Broad S.ωÞfJe Management

Acco削1ingSyslem, Job lnformation Perceplions, and Managerial Peiformance

Figure

3

Path Diagram for

All

Samples

P21 = P32 = Broad Scope 0.310** P1RJ (z,) 。 443* ﹒ M阻agerial MAS (Z,) Perfonnance (Z,)

已

Rv P31=0.103 L一一 Note:**p<O.OI

Later, we decomposed the correlation coefficient r;j using the following equations into direct effect, indirect effect or spurious effect

rl2 = P21 rl3 = P31

+

P32 rl2 r23 = P32

+

P31 rl2 (3) (4) (5)

In the above Equation (3), the value of P21 is equal to the correlation coefficient r12. This means if one variable has only one antecedent variable, the path coe旺icient between the tWo variables is the co汀elation coefficient between the two variables. In Equation (4) 創d Equation (5), the first tenn on the right side of the equations， 恥， is the direct effe仗， while the se∞nd tenn P;jT;j denotes the indirect effect or spurious e缸èct. Equation (4) decomposes the correlation coefficient between broad scope MAS and managerial perfonnance r13 into direct effect P31 and the indirect effect P32r12. Equation (5) decomposes the correlation coefficient between PJRI and managerial perfonnance r23 into direct effect P32個d spurious effect P31r12. This spurious effect is induced by the common antecedent

Chìao Da Managemenl Review 均1. 29 No. 2, 2009 65

variable of P JRJ and managerial performance- broad scope 恥的S

The decomposition results ofEquation (3), (4), and (5) with all samples are presented in the second - fourth columns ofTable 5. The results ofEquation (4)

reveal that the total effect of broad scope MAS on managerial performance (0.241) can be decomposed into direct effect (0.103) and indirect effect ofPJRJ (0.138)。 The indirect effect is greater than the direct effect. According to Baron and Kenny

(1986), for a mediator variable between an independent variable and a dependent

variable to be suppo巾d， three conditions must be satisfied: (1) the independent

variable has a significant effect on the mediator variable; (2) the independent

variable has a significant effect on the dependent variable; and (3) after the dependent variable is regressed on both the independent and mediator variables, the mediator variable has a significant effect on the dependent variable, and the regression coefficient of the independent variable over dependent variable is lower than the regression coefficient under Condition (2). Morωver， m ∞ndition (3), if the regression coefficient of the independent variable over the dependent

variable is also significant, the mediator has only partial mediating effect; if n仗， the mediator has full mediating effect.

It can be inferred from Equation (3) that when a variable has only one antecedent variable. the correlation coe伍 cient of the two variables is the

regre泌的n coefficient of the two variables. As shown in Table 3, the relationship

between broad scope MAS (independent variable) and 的 RI (mediator variable) is

significant, and that between broad scope MAS (independent variable) and managerial performance (dependent variable) is also significant. Condition (1)

and Condition (2) of mediation proposed by Baron and Kenny (1986) are satisfied. Through regression analysis of Equation (2) and the results shown in Table 4, we find that when managerial performance is regressed on both broad scope MAS

and PJRI, the relationship between PJRI and managerial perforrnance is

significant (P32 = 0.443, P < 0.01), and the regression coefficient of broad s∞pe

MAS over managerial perforrnance (P31 = 0.103, P > 0.1) is smaller than the regression coefficient (0.241) derived in Condition (2). Thus, Condition (3) is also

satisfied. Besides, under Condition (3), the regression coefficient of broad s∞pe MAS over managerial performance is not significant. This reveals that mediation

66 The Impacl of Slralegy 砂'Peon the Relationships among Broad Scope Management

Accow的 ngSyslem, Job Informalion Perceplions, and Managerial Performance 。f P lRl in the relationship between broad scope MAS and manager凶 perfo口nance is supported, and P JRl has full mediation effect. Thus, H 1 is supported. This finding is consistent with the ∞nclusions of Kren (1992) and William and Seaman (2002) that the relationship between management accounting systems and performanc唱的 mediated by perceptions ofjob-relevant or managerial relevant information

Table 5

Decom position of the Path Analysis Model

Links A1 1 阻mples DifTerentiation strategy Low-∞5t 5個tegy

(n ~ 107) (n ~ 54) (n ~ 53)

Total Direct lndirect Total Direct lndirect Total Dif'CCt lnd甘eot

effect effect or cffcct e叮ect 。r e何ect efT，目t or

Spurious Spurious Spurious

effect e叮ect effect

Equation (3) PJRl!BSMA 0.310. 0.310.. 0,474.. 0.474.. 0.184 0.184 s

•

Equation (4) Managerial perfonn扭曲/ BSMAS 0.241

.

0.103 0.138" 0.350" 0.124 0.226' 0.132 0.056 0.076" Equation (5) Managerial pcrfonn 扭扭/ PJRJ _。‘475. 0.443.. O.032b 0.535.. 0_.477" O.05Sb 0_.425.. 0_.415.. _O.OIOb •

Note: • indirect effect. b spurious effect. BSMAS = broad s∞pe MAS. P1RJ ~ perception ofjob-relevant

infonnatior>

..p < 0.01. .p < 0.05.

4.2 The Inf1uence of Strategy Type on Relationships among Focus Variables As shown in Table 3, strategy type is positively ∞ηelated with broad scope MAS but not to a signi日cant extent (r = 0.068, P > 0.1). This manifests that strategy type does not influence the level to which managers use broad scope MAS. Thus,

H_2a is not supported, This finding is ∞nsistent with the ∞nclusion of Baines

Chiao Da Managemel11 Review Vol. 29 No. 2, 2009 67

when enterprises tend to employ the differentiation strategy, they will rely more on non-financial management accounting information. However, such hypothesis was not empirically supported

Table 6

Correlation Matrix (Differentiation Strategy

,

n = 54)

Measurement Variable Broad Scope MAS PJRI Managerial PJRI Managerial Performance Strategy Type Note:**p < 0.01 。 474** 0.350** -0.068 Table 7 0.535** 。 106

Correlation Matrix (Low-Cost Strategy

,

n = 53) Measurement Variable PJRI Manageriai Performance Strategy Type Note:** P < 0.01. *p < 0.05 Broad S∞peMAS 。 184 0.132 0.216 PJRI 0 .4 25 拿， 。 284* Performance -0.079 Managerial Performance 0.057

To test the effect of strategy type on the relationship between broad scope MAS and 的悶， and the relationship between P JRl and managerial performance,

we divided the samples by median into two groups: differentiation strategy (n =

54) and low-cost s甘ategy (n = 53). The correlation matrices ofthe two groups are presented in Table 6 and Table 7. We further conducted path analysis to estimate

pa甘1 ∞e缸icients p;j. The results are shown in the fifth and sixth columns (differentiation strategy) and the seventh and eigh伽 columns (Iow-cost s甘ategy) 。fTable 4, and shown in Figure 4 and Figure 5.

Like for a11 samples, the direct li成 be仰veen broad s∞pe MAS and managerial performance (P3t) is insignificant for either group of different str帥gy. Th的 indicates that there is no correlation between broad s∞pe MAS and

的 The 1m.戶口 ofSI間legy 砂'pe on the Relationships among Broad Scope Managem帥，

Accounling Syslem, Job 1nfom1Glion Pe間plions， and Managerial Performance

managerial performance no matter which strategy type is employed. On the other

hand, the path coefficient between broad scope MAS and P JRI (PZI) becomes extremely significant (0.474, p < 0.01) when an ente叩 rise tends to employ the differentiation strategy. This result indicates that broad scope MAS is positively

correlated with P J則 If the enterprise tends to employ the low-cost strate釘， PZI

becomes insignificant (0.184, p > 0.1), meaning that there is no ∞付elation

between broad scope MAS and PJRI. lt can be infe叮ed that when an ente巾nse

tends to adopt the differentiation s甘ateg)心 there will be a stronger relationship between broad scope MAS and PJRJ. The result ofthe generallinear tese (Neter

& Wasserman, 1974) shows a significant di仔erence between path coefficients (PZI)

under the two strategy typesσ'Z，103 = 2.530 > _FZ,I03,O.1 = 2.347). Therefo時， HZb is

supported. This finding is consistent with the argument of Williams and Seaman

(2002) that when task difficulty is high, changes in management accounting and

control system can induce a higher level of per∞ption of managerial relevant

information.

However, the analysis result does not support H_zc. As shown in Table 4, the path coefficient between P JRI and managerial performance (P32) 的 greater under di能renti的 ion strategy (0.477, p < 0.01) than under low-cost strategy (0.415, p <

0.01). Under both s甘ategies， PJRI is significantly correlated with managerial performance. Besid郎， the result of general linear test shows no significant

difference between the two (F門，IOI = 0.593 < F₃,101,O.1 = 2.130). This implies that

the positive relationship between P JRJ and managerial perforrnance cannot be

augmented through adoption öf 的e differentiation strategy. As a result, H_zc is not

empirically supported. This finding is consistent with the conclusion of Kren

(1992) that whether there will be a stronger relationship between PlRI and

managerial performance under a high level of environmental uncertainty cannot

7 General lin開rlesl is usually applied to test 出e correlation between an independent variable and

a dependent variable is significant different in two groups of data, one hi吵 and one low. ln the

test of the difference betw自n 出e regression models of the two grou戶，the estimate (Fηderived und冒出e numerator's degree of freedom K + I and denominator's degree of 告eedom

n+M-2(K+I) is ∞mpared wi曲曲e lest value (F) (wh叮e K denotes the number of regression coe伍cients to be estima閥 n 阻d M are respectively the sample size ofthe two groups). IfP IS gr閏即出an F, there is significant di能rence between the two regression models (Neter &

Chiao Da Management Review 均1. 29 No. 2, 2009 69

be verified. Besides, Will iams and Seaman (2002) also mentioned that whether

the relationship betw巴en perception of managerial relevant inform泌的n and

departmental performance can be stronger under high task uncertainty cannot be validated.

The decomposition results of Equation (3) - (5) are shown in the fifth

-seventh columns (differentiation strategy) and eighth - tenth columns (low-cost

strategy) of Table 5. As indicated by the results of Equation (4), for both groups

of enterprises, broad s∞pe MAS has a stronger indirect effect (through PJRI) than

direct effect on managerial performance. Besid郎， the indirect effect under the

di前erentiation strategy (0.226) is greater than the indirect effect under low-cost strategy (0.076). This is the expected result ofthis study: when an enterprise tends

to employ the differentiation strate缸， a higher degree of use of broad scope MAS

can lead to a higher level of P JRI and hence improved managerial performan∞

The correlation matrix in Table 6 shows that under differentiation strategy,

broad scope MAS has a significant e前ect on P lRI and also managerial

performance. Hence, Conditions (1) and (2) of mediation proposed by Baron and

Kenny (1986) are met. Besides, as shown in the fifth and sixth columns ofTable 4,

after managerial performance is regressed on both broad scope MAS

and PJ肥， PJRI has a significant effect on managerial performance (P32 = 0.477, P

< 0.01). The regression coefficient of broad scope MAS over managerial

performance (P31 = 0.124, P > 0.1) 的 smaller than the regre泌的n coefficient in

Condition (2) (0.350), and under the level of significant. According to Baron and

Kenny (1986), PJRI has full mediating effect. Through a ∞mparison between

Figure 3 and Figure 4, we found that the path coe叮icient between broad scope

MAS and P1RI (P21 = 0.474) and the palh coellicient belween PJRI and

managerial performance (P32 = 0.477) obtained from enterprises using the

differentiation strategy are greater th個 the path coefficients obtained from all

samples (P21 = 0.310, P32 = 0.443). Therefore, when an ente叩rise tends ωadopt

the differentiation strategy, the full mediating effect of P1RI on the relationship

between broad scope MSA and managerial performance will become more pronounced

70 The Jmpacl 0/ Slralegy Type on Ihe Relalionships among Broad Scope Managemenl

Accounting 秒'stem， Job lnformation Perceplions, and Managerial Performance

As shown in Table 7，山lder the low-cost strategy, the effect of broad scope

MAS on either PJRJ or managerial performance is not significant. The conditions

of mediation proposed by Baron and Kenny (1986) cannot be satisfied, so

mediation of P J間的 not supported. This a1so indicates that when an ente叩nse

tends to employ the low-cost strategy, high level of use of broad scope MAS

cannot enhance PIRl and managerial performance.

Figure

4

Path

Di

agram for Samples of

Di

fferentiation Strategy

P21 = P32 = Broad Scope 0.474" PJ閱 (Z，) 0.477" Managerial MAS (Z,) Perfonn個閱 (Z，)

•

之

θ

Rv P3I = 0.124 Note: "p<O.O 1 Figure 5

Path

Di

agram for Samples of Low-Cost Strategy

P21 = p" = Broad Scope 。 184 PJRl (z,) 0.415" M個 agerial MAS (Z,) Perfonnance (Z,)

θ

之

Rv P31 = 0.056 Note:"p<O.OI

Chìao Da Management Revìew 均1. 29 No. 2, 2009 71

ln sum, for a11 samples or those inclined to employ the differentiation s甘ategy， the mediation of PJRJ is supported. However, the effect of strategy type on the relationships among variables is only partially supported. That is to s旬， for enterprises inclined to adopt the differentiation s甘ategy， there is a stronger correlation between broad scope MAS and P J悶， but the level of using of broad

scope MAS is not necessarily higher, and the correlation between P JRJ and managerial performance is not stronger either. The reason may be the fact that in faωofunce此ainty of the extemal environment, managers need broad scope MAS

to provide extemal, non-financial, and fu仙re-oriented informati凹， so the level of using broad scope MAS is not affected by the strategy type adopted by enterprises

However, when enterprises are inclined to take the differentiation strategy, they

need broad scope management accounting information, including customer

preferen臼 and technology development. Therefore, using broad s∞ pe MAS at a

high level can lead to a high level of PJRJ. ln additi凹， the relationship

between PJRJ and managerial performance is extremely significant under either

type of strate目的， implying that managers consider job-relevant informati凹的

critical for accomplishing goals of their jobs. So, no matter which strategy is

adopted, P1R1 of managers has absolutely positive effect on their managerial

performance

5. Conclusions and Limitations

ln 曲的 study， we probed into the relationship between broad s∞pe MAS

and managerial performance using PJRJ as a mediator variable. Expected results

were obtained. The relationship between broad s∞pe MAS and managerial

performance is indeed mediated by PJRJ. That is to say the positive effect of

broad s∞pe MAS on managerial performance is attributed to an increased level

of PJRJ. Bes泌的， P JRJ has full mediating effect on this relationship, and this

effect becomes more pronounced when enteφrises are inclined to employ the

72 The lmpact ofStrategy砂peOn the RelofÎOnships among 8road Scope Managemenl Accounting Syst酬， Job lnformation Pe陀eptions， and Managerial Performance

Theoretically, our inclusions of moderator and mediator variables in 出e

discussion of the relationship between broad scope MAS and managerial

perfo口nance better explained the relationship in a more complete model. In

addition to the findings from the perspective of the contingency theory, there is

full mediation of P1RI, which bec心mes s甘onger when the differentiation strategy

is adopted. Practically, the following managerial implications were provided: The

effect of broad scope MAS on managerial performance is supported only when

managers' PJRI can be induced. When enterprises are inclined to adopt the differentiation s甘ate釘， broad scope MAS should be adopted

at

the same time, so

as to enhance managers' PJRI and performance. lf the low-cost s甘ategy is

adopted, managers are able to accomplish tasks without much additional

information. In this case, it is not necessary to use broad scope MAS. Adoption of

broad s∞pe MAS may only cause information overload and waste of i的∞sl. lt

has no positive effect on the managers' P 1RI and performanω1 n addition, the positive correlation between P1RI and managerial performance does not vary

under different strategies. Whether the differentiation strategy or the low-∞st

strategy is adopted, efforts should be made to enhance managers' PJRI to further

improve the甘 perfo口nance. AII these findings can help ente中rises make a better

use ofbroad scope MAS.

Future researchers are suggested to explore whether the mediating effect

of P1RI on the relationship between broad s∞pe MAS and managerial

performance would vary in di位rent contexts (such as a decentralized

organizational s甘ucture and different information processing abilities). ln this

study, the plant manager's managerial performance was not evaluated by both the

manager's superior and subordinate. Possible inconsistency between the

superior's and the subordinate's evaluations was not considered. Therefore, future

research闊的 advised to let a manager's managerial performance evaluated by

both the superior and subordinate when perception measurement is used

Questionnaire survey method was adopted in this study, so there would be

some generallimitations ofthis method. Take the items designed to measure P1RI

Chiao Da Management Review 均1. 29 No. 2, 2009 73

。 7 standard, so the first item8 with a larger measurement error was not deleted. This was one of the limitations of this study. Besides, the subjec尬。fthis research were plant managers and their subordinates of electronic firms in Taiwan. Thus,

the research results cannot be generalized and applied to different departments,

industries or companies in other nations

Appendix 1

8asic Data ofThe Respondents

Demographic Group Plant Manager Subordinate

Variable

Age 23-29 Years Old 2 1.9% 29 27.1%

30-39 Years Old 18 16.8年也 43 的 2% 40-49 Years Old 57 53 .3早色 _{22 20.6%} 50-59 Years Old 25 23.4% 6 5.6% 60-69 Years Old 0.9% 4 3.7% N/A 4 3.7~也 3 2.8% T。個1 107 100.0% 107 100.0% Seniority 0-5 Years 83 77.6% 80 74.8% 6-10 Years 17 15.9~也 18 16.8% l 卜的 Years _{0.9% 6 5.6%} N/A 6 5.6<}也 _{3 2.8%} Total 107 100.0哥也 107 100.0% Education Graduate Institute 11 10.3年也 12 11.2%

Background _{College/University 81 75.7年色 83 77.6%}

Senior

High/V ocational 10 9.3年色 7 6.5年也 School

Junior High School

_。

_。

2 1.9%

N/A 5 4.7% 3 2.8%

To旭l 107 100.0% 107 100.0%

8 甘le measurement error m 出is item could be caused by respondents at甘冒 but自由eir perform聞自

74 The Impacl ofSlralegy 砂'Peon the Relationships among Broad Scope Management

Accounting 丘ySlem， Job Information Perceplions, and Managerial Performance

Broad Scope MAS

Appendix 2 Questionnaire

Please evaluate the degree to which you use the following inforrnation in your work. Put

in box to the box to indicate your answer

I.Inforrnation related to

possible future events

(e.g. new legislation)

2.Quantification ofthe

likelihood offuture

events occurring (e.g.

probability 的timates) 3.Non-e∞nomlc information (e.g customer preferences, employee attitudes,

labor relations, attitudes

of govemment and

consumer bodies,

competitive threa臼，

etc.)...

4. Information on broad

factors extemal to your

organization (e.g

economlc ∞nditions，

population growth,

Never Less Seldom Sometimes More 曲曲 Frequently A1ways

than Sometimes

Seldom

口 口 口 口 口 口 口

口 口 口 口 口 口 口

Chìao Da Management Revìew 均1. 29 No. 2, 2009

technological

developments,

etc.).. ..

5. Non-financial

infonnation that relates to production

infonnation (e.g. output

rates, scrap levels,

machine efficiency,

employee absenteeism, etc.).. .

6. Non-financial

infonnation that relates to market information (e.g. market size, growth share, etc.).. . 口 口 口 口 口 口 口 口 口 75 口 口 口 口 口 口 口 口 口 口 口 口

76 The lmpact ofStrategy Type on the Relationships among Broad Scope Management

Accounting System, Job lnformation Perceptions, and Managerial Pe廿ormance

Perception of Job Relevant /，呵{ormoJion (PJRl)

Please evaluate the degree to which you agree each ofthe following questions? Put

a check in the box to indicate your answer.

S釘ongJ y Very Very Strongly

Disagree Neutral Agree

Disagree Disagree Agr田 Agree

1. You are always clear about what is

necessary for you to perfonn wel1

。nyo區 job. 口 口 口 口 口 口 口

2. You have adequate inforrnation 10

make optimal decisions to

accomplish yo叮 perforrnance

。阿ectlves 口 口 口 口 口 口 口

3. You are ahle to ob旭in the strategic

information n聞自阻ryto ev剖uate

Import血t deci日 onalternatives 口 口 口 口 口 口 口

Strategy 乃少e

Please evaluate the percentage oftotal tumover ofyour plant (in 100%) gained 宜。m

the differentiation strategy and the percentage from the low-co咀t strategy? (Please

note the explanati叩 below)

Differentiatio叫 S甘ategy: ( )自

Low-∞st S甘ategy: ()姑

100克

Explanation:

Differentiation Strategy: The main goal of this strategy is to create certain product

features. S甘ategies such as working out outstanding designs,

providing good customer service, establishing brand

awareness, and producing high-quality products can be

classified as differentiation strategy

Low-cost s甘ategy﹒ The main goal of this strategy is to reduce product cost to the

Chiao Da Mal1agemenr Review Vol. 29 No. 2, 2009 77

Managerial Performance

Please evaluate the performance of your plant manager in comparison with

executives of other departments in the following aspects. Put a check in the box to indicate your answer

Awful Very Poor Acceptable Good Very Excellent

l. Goal and task scheduling... .. 口 2. Info口nation processing and task

research 口

3. Communication and coordination

WI曲。由erdepartments 口

4. Appraisal ofpersonnel aod 旭sk

relevant items 口 5. Supervision over personnel and task

re]evant items 口

6. Employment of personnel... ... .. 。 7. Negotiation with extemal

organl甜tions(such as suppliers and

customers) 口

8.lncτease of the organizational

tnterest 口

9. Overall performance 口

Poor 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口

6. References

Good 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口 口

Abemethy, M. A. and Brownell, P. (1997), “Management Control Systems in Research and Development Organizations: The Role of Accounting,

Behavior and Personnel Controls," Accounti嗯， Organizations and Socie紗，

78 The lmpacf ojSfrafegy 砂'p"on fhe Relafionships among Broad Scope Management

Accounfing Sysfem, Job lnjormafion Percepfions, and Managerial Performance

Anderson, 1. C. and Gerbi嗯， D, W, (1988), “Structura1 Equation Modeling

in Practice: A Review and Recommended Two-Step

Approach," Psychological Bulletin, 103(3), 411-423

Asher, R, (1983), Causal Modeling, London: Sage

8aim姐， S. (1982), “Agency Research in Managerial Accounting: A Survey,"

Journal of Accounting Literature, 1, 154-213

8aines, A.個d Langfield-Smith, K. (2003),“Antecedents 10 Management

Accounting Change: A Structural Equation Approach," Accounting,

Organizations and Socie紗， 28(7/8)， 675-698

8ank缸， R. D. 8ardhan, 1. R. and Chen, T,y' (2008), “The Role of

Manufacturing Practices in Mediating the Impact of Activity-8ased Costing

on Plant Performance ," Accounti吟， Organizations and Society, 33(1),

1-19

8ar凹， R. M. and Kenny, D, A. (1986), “The Moderator-Mediator Variable

Distinction in Social Psychological Research: Conceptual, Strategic and

Statistical Considerations," Journal of Personality and Social Psychology,

51(6), 1173-1182

80uwens, ], and Abemethy, M. A. (2000),‘'The Consequences of Customization

on Management Accounting System Design," Accounting, Organizations

and Society, 25(3), 221-241,

8rownell. P. and McJnn郎， M. (1986),“8udgetary Participation, Motivati凹， and

Managerial Performance," The Accounting Review, 61(4), 587-600,

Chalos, p, and Poon, M. C. C. (2000),“Participation and Performance in Capital

8udgeting Te街ns，" 8ehavioral Research in Accounting, 12, 199-229

Chandler, A.

0

, Jr. (1962), Strategy and Structure, Cambridge, MA: f'，位T Press,

Ch個皂， R. D. and Chang, Y. W. (2001),“The Effects of Strate釘， Organizational S甘uct叮e and Environment on 血e Characteristics and Performance of

Management Accounting 1nformation Systems," Management Review,

20(4), 137-173

Chan皂， R. D. Chan皂，

y.

W, and Ni, f , y , (2000),‘'The Effect of Task Attributes and AlS Characteristics on Users' Satisfaction: An Empirical Test,"

Chiao Da Management Review Vol. 29 No. 2. 2009 79

Che仙all ， R. H. (2003), “Management Control Systems Design Within Its

Organizational Context: Findings from Contingency-Based Research and

Directions for the Future," Accounting, Organizations and Socie紗， 28(2/3),

127-168

Che叫lall ， R. H. and Brownell, P. (I988),‘'The Effects of Participative Budgeting

on Job Satisfaction and Performance: Role Ambiguity as an Intervening Variable," Accounting, Organizations and Socie紗" 13(3),225-233

Che叫lall ， R. H. and Morris, D. (1986),“The Impact of Structure, Environme肘，

and Interdependence on 也e Perceived U sefulness of Management

Accounting Systems," The Accounting Reνiew， 61(1), 16-35

Chi

a,

Y. M. (1995),“Decentralization, Management Accounting Systems (MAS)

Information Characteristics and Their Interaction Effects on

Managerial Performance: A Singapore Study," Journal

01

Business Finance and Accounting, 22(6), 811-830.

Chong, M. L. and Edmund, W. L. (2002), ‘'The Intervening Effects

。f Participation on the Relationship between Procedural Justice and

Managerial Performan氓" British Accounting Review, 34(1), 55-78

Chon皂， V. K. (1996),“Management Accounting Systems, Task Uncertainty and

Managerial Performance: A Research Note," Accounti嗯; Organizations

and Society, 21(5), 415-421

Chon皂， V. K. (1998), “Testing the Contingency “自t" Between Management

Accounting Systems and Managerial Performance: A Research Note on the

Moderating Role of Tolerance for Ambiguity," British Accounting Review, 30(4), 331-342

Chong, V. K. and Chon皂， K. M. (1997), “Strategic Choic郎， Environmental

Uncertainty and SBU Performance: A Note on the Intervening Role of

Management Accounting Systems," Accounting and Business Research,

27(4),268-276

Cronbach, L. J. (1951 ), “Coefficient Alpha and the Intemal Structure of Tests," Psychometri切， 16(3), 297-334.

DeGroot, T. and Brownle巴， A. L. (2006),“Effect of Department Structure on the