The purpose of this chapter is to review the various streams of audit structure, audit JDM, and audit risk literature that have a direct relation to this dissertation's research questions.
The first section describes the SSA audit structure in depth and identifies key differences between this approach and the TBA structure, differences that may affect auditors' judgment on the client risk. The second section details the determinants of the auditor Judgment decision making "JDM" process that built from applying one of these two structures and review the main finding which are relevant to this study, namely, assessment of inherent risk and control risk . The third section reviews the types of business risk that could surround the clients business namely, external and internal business risks. The final section reviews the studies that discusses the relationship between the three component of this study namely, the relationship between audit structure and audit risk, the relationship between audit structure and "audit "JDM", and the relationship between audit risk and "JDM". The key findings from these streams of literature will then be assimilated in Chapter 3, where the dissertation's hypotheses are developed.
Audit structure
The late 1970s and 1980s shows a changes in the US market rules; the deregulation of the economic systems opened new opportunities to US corporations to expand their activities around the world. As a result, large accounting firms (the big 8) traced their customers under the incentive of expanding their services and seeking benefits from the deregulation of the market (Baskerville 2010). The change in the competitive conditions for the accounting firms was another major event led by the removal of most advertising restrictions in 1978, and the direct uninvited solicitation restrictions during the same period, (Chaney et al. 1995). This increased the competition between accounting firms. As a response, the audit fees were under pressure to be reduced ,because clients can shop around for the audit service (Humphrey 2001). This indicates the highly competitive environment during the 1980's that led audit firms to compete through cutting the audit cost. Additionally, the rapid growth of the audit procedures and practices, the expansion of the personnel, improvements in technology and the exigencies of reducing costs, accounting firms began to design and implement highly structured processes.
It can be notes that this stream starts from observing the audit firms strategies when performing an audit judgment about the client's financial activities where the large audit firms shift their process such as (refinements of statistical sample, risk detection tests, analytical procedures, decision-making aids, and going concern evaluations) to more structured audit in order to enhance the audit efficiency and maintain a high level of audit quality. The term "structure" become popular during the 1980's .
Snowball & Tubbs (1989, p. 286) describe structure as "the arrangement of people, tasks, and authority to achieve more calculable and predictable control over organizational performance. One of the primer studies that starts this debates formed by Sullivan (1984) who argues that auditors are following two process when perform the audit. He argues that two camps are exist . The first camp prefer a structured quantitative algorithms over auditor judgement. The second camp prefer professional judgement because auditor need to perform significant amount of judgment over their clients' accounts. This indicates the major different between structure and unstructured audit process. Moreover, This debates motivates researchers to investigates which method can achieve better results. For example, study by Kinney (1986) suggest a classification of accounting firms as "structured," "intermediate" or "unstructured" in terms of their audit methodologies. He apply a survey over 22 audit firms in order to classify them according to his suggestion . 18% of audit firms were found to apply a structured audit, another 18% audit firms found under intermediate structures and 64% using unstructured audit mostly from non-big audit firms Additionally, he Argues that those firms with a structured audit approach will manage audit risk through a greater reliance on mechanistic procedures, resulting in a greater tolerance of income manipulation.
Morris and Nichols (1988) show that firms applying structured audit are more consistence with accounting principle than firms with other approaches . his study starts the investigation about which method can generates better results for both audit companies and community.
As mentioned before, the debates between the two camps has motivated the researchers to investigate this phenomena, a study by (Dirsmith and Haskins, 1991). Suggest to form a basis to refined the reasoning process between the two method (structured and unstructured) as they notes that both control and inherent risk interacts with each other and must to consider during the audit planning stage in order to determine the detection risk. Therefore, a structured audit may impact the evaluation of the inherent risk . and more thorough evidence evaluation is need . .They suggest two theories that researchers can use to assess the audit risk "mechanistic world" and "organic world" hypotheses.
The first theory "mechanistic world" consider a structured audit process which concentrates on segments, priority relationship with these segments and the dominant side of quantitative versus qualitative components when execute an audit judgment. The second method consider the audit as a judgment process which concentrate on building a holistic integration with more focusing on qualitative process.
.
(Libby 1985) shows that structured task audit may help the auditors to linked information acquired from the audited with mental representation. This link may increase auditor ability to identify inconsistencies during analytical procedures(ex: evaluates pattern of account balance changes), and internal control evaluation. He argues that a well formed mental representation derives from the auditor built up knowledge. His argument motivated other researchers to investigate the relation between the auditor experience, knowledge, with the mental representations. -This part will be discuss in later section of this dissertation-. Moreover, a later study by (Libby and Luft, 1993) focused on identifying the interactions between determinants of performance, specifying underlying cognitive processes, and abstraction based on theory and task analysis.. the study reveals that the task characteristics influence and affect the audit judgment
(Bonner & Walker 1994) argues that the presence of task structure for the auditor should influence the auditor judgment about the diagnostic reasoning for the evidence gathered from the client's business. Moreover, task structure prepared by the audit firm influences the auditor's evaluation of the evidence during analytical procedures (Biggs et al. 1995). Also, some other researchers have provided evidence about the influence of task structure on auditor's knowledge and cognitive learning in order to complete a complex judgment task (Zimbleman, 1997, Kachelmeier and Messier, 1990) Also, task structure techniques has overcome the information overload through performing a complex task( Kleinmuntz, 1990 Check Edaarah), Hence, many studies had confirmed a positive relation between the effectiveness of auditor's performance and the existence of audit structure when dealing with complex task (Butler 1985; Kachelmeier and Messier 1990; McDaniel 1990; Messier et al. 2001; Zimbleman 1997). The previous literature suggests a direct relation between the existence of audit structures and the effectiveness of the audit.
.Furthermore, (Salterio, Knechel 2007, and Kotchetova 2006) have provided further evidence about audit firms with different task structures could influence the results of the auditor's risk assessment. Additionally, Kennedy (1995) found that the manipulations of the task structure and the effort that the auditor used in analysing pattern of change accounts affect the auditor's judgment about misstatement risk that arise from the assessment of the client's business risk .Therefore, we can conclude that the existence of task structure affect the audit results from different aspects, they are as follow:
1- Influence the auditor ability to diagnose the collected evidence
2- Influence the auditor cognitive learning to complete a complex judgment task
3- Different audit task structures generates different audit results
4- Influence the auditor judgment about misstatement risk that arise from the assessment of the client's business risk
5- Influence the auditor performance when dealing with complex task
From the above we can form a conclusion that audit firms who use an audit structure technique consider both qualitative and quantitative risk factors when performing the audit plan. The auditors quantitatively integrate evidence based on systematic process that planned by the accounting firms and qualitatively build a mental representation which aims to retrieve and linked the obtained information. On other words - structured audit provide a formed steps to able the auditor to enhance the cognitive process in order to facilitate information integrations, On the other hand an unstructured audit which emphasis on qualitative process relies on holistic integration for the audit evidence that may increase the burden on auditors memory.
The next part of this section review the literature of two known audit structured that covers the scope of this dissertation. The first structured refers by researchers as the "Bottom-Up" audit structure which derives from traditional audit technique and led by many researchers during the 1980's and the beginning of 1990's such as (Koonce 1993, Libby 1985,Libby and Luft 1993). The second audit structured is referred to by researchers as the "Top-Down" audit structure that derives from "SSA" audit, which has become very famous term in the audit literature during the late 1990's and has been advocated by many researchers such as Bell (1997), Knechel (2007), Lemon (2000) and Ketchecova (2003).
JDM in audit
Behavioural decision theory became very popular in the accounting profession during the early 1980s (Libby 1985), for instance, Libby and Lewis (1982) cited 61 accounting articles in this area. Ponemon & Gabhart (1990) argue that most of these studies "focused on the impact of decision makers' in individuals and small groups basic cognitive abilities and deficiencies on decision performance". Accounting literatures determined the decision performance factors by studying other social science fields, such as psychology). Accounting researchers classified the factors that affect decision performance in accounting as follows:
Information process
Short memory V.S long memory
Selective attention
Lack of cognitive control
Cost of thinking
(Libby and Lewis 1982) argue that accounting field was still premature at this stage. As it struggles to capture much of the richness of accounting decision settings.
The mid 1980s faces a dramatic change in decision-making process researches. As the literature started to distinguish the decision making process in accounting from psychology (Ponemon & Gabhart, 1990). This separation took the researchers toward two trends of "JDM" studies in accounting as follows:
"JDM" focus in task-knowledge
"JDM" focus in the environmental effect
The first trend was interested in investigating the role of the individual experience related performance and knowledge differences, where the auditor experience was the main factor that distinguishes the individual knowledge. As a result, many studies found that experienced individual can performer a better decision-making performance than non-experienced ones. This led researchers to investigate the knowledge's effect on the proper use of "JDM" paradigm in one hand , and the effect of auditors knowledge differences in using memory tasks such as (knowledge storage and knowledge organization) on the other.
The second trend focused on the role of environmental effect on "JDM" which concentrated on more distinctive characteristics of accounting settings such as :the role of conflict and ambiguity on audit structure (Bamber et al. 1989), uncertain task structure (Bamber et al. 1988), the effect of accountant decision performance under time pressure (McDaniel 1988).
Audit profession is based on performing a professional judgment through directed goal process of deciding and/or choosing of possible actions that are subjective to due care, objectivity, and integrity within an appropriate framework of professional standards (Gibbins & Mason, 1988). (Ponemon & Gabhart, 1992) argue that the nature of audit environment is well suited for studying professional judgement in audit. The next millennium shows a rapid growth of behavioural studies on audit. A recent research by (Nelson and Tan 2005) reviews the "JDM" in audit for the last three decades. They suggest organizing audit researches for (JDM) under certain categories as follow:
1- JDM in audit task
2- JDM on the auditor attributes
3- JDM on interpersonal interaction
JDM in audit task researches focused on the effect of audit task in decision-making, (Nelson and Tan 2005) categorized the literature of JDM in audit task into the following:
1- Risk assessments (including the audit-risk model and related audit planning decisions)
2- Analytical procedures and evidence evaluation
3-Auditors' "correction decisions" regarding whether to require clients to book proposed adjusting journal entries
4- Going concern judgments
The second category of (Nelson and Tan 2005) are the studies about JDM on the auditor attributes which focuses on auditor responds to any events during the audit , these studies are involved in auditors attitude and behaviours toward those events .The NT's categorize those studies as follows:
1- Auditor knowledge and expertise ability
2- Individual Characteristics
3- Cognitive Limitations
4- Decision Aids
The third and the last category of (Nelson and Tan 2005) are studies about JDM on auditor interpersonal interaction. These studies cover wild varieties of auditor's interaction between each other, clients, and communities. NT's categorized these studies as the following:
1- Between-Auditor Interactions
2- Auditor/Client Interactions
3- Interactions between Auditors and Others
From the above, it can be noted that the early attempts to investigate the "JDM" in accounting as a social science had set the ground for later development in this field. The development starts by adopting other social science methods and dissolve them within the accounting profession. This adoption failed to explain "JDM" process due to the unique accounting settings from other social science settings. Later on, researches started to be distinguished accounting process from other sciences, which resulted in a unique framework that organized "JDM" in accounting and later on the audit profession. The appendix () shows the researches progress in this field until the dissertation time.
(Nelson and Tan 2005) attempt to organize "JDM" researches for the last three decades in three categories mentioned above. However, avoiding the interaction between these categories is nearly impossible as the "JDM" process in audit task cannot be performed without the auditor forming a mental representation for the client. Therefore, analysing the audit task itself will ultimately lead to an interaction between the audit tasks with auditor's attributes. However, following a well-constructed model may eliminate some of the auditor's attribute effects.
The relationship between audit task structure and "JDM"
According to Simon (1978) model the decision-making process encompasses with three repetitive phases, they are:
Intelligence
Design
Choice
At intelligence phase, the auditor defines and diagnoses the problem in order to gather related data to the problem and then validate it. At the design phase, the auditor generates alternative set of actions in order to assign the task. The auditor may also extend the search of data if necessary. The third phase involves explaining the alternatives by generating a statistic relation for the task in order to choice the best alternative.
(Keen & Scott-Morton 1978) argue that (Simon 1978) decision-making model varies between audit tasks. They suggest that in some tasks, the problem can be well defined at the intelligence phase, which leaves little to do in the design phase. As a result, choosing between alternatives will be very limited. In other words, the audit judgment will be very restricted for the task. They call this form of task a "programmable" or "structured" task, as well as a "nonprogrammable" or "unstructured" task identified when it is difficult to define the problem at the intelligent phase which leads to a large number of alternatives during the design phase. In this case, a considerable judgment and insight is required during the choice phase. Between the two extremes a "semi programmable" or "semi structured" task relies on a reasonable definition to the problem during the intelligent phase, a limited and specified number of alternatives are indentified during the design phase which will consequently lead to some judgment that will be exercised during the choice phase.
(Simon 1960) model set the ground for further investigations for the relationship between "JDM" and the task structure, As well as ( Bonner 1991 , Libby 1995) emphasized the need to consider the task structure in behavioural researches (refer to JDM in audit literatures) . As the audit processes contain a numerous number of tasks, the nature of audit judgment is expected to vary between these tasks according to their nature. (Keen & Scott-Morton 1978) classified these tasks in order to understand the amount of judgment needed to be exercised for each task and to understand the expected outcome of these judgments. (Cushing & Lebbecke, 1983) extended (Keen & Scott-Morton 1978) tasks inventory by indentifying 28 audit procedures to assess the audit methodology in 14 audit firms. (Chow et al. 1987) extended these procedures into 60 audit steps to investigate the difficulty and the significance of these steps. In a more recent research, Abdolmohammadi, (1999) relied on 3 big audit firms, and a study of 14 large audit firms by Cushing & Lebbecke (1983). He provided a comprehensive inventory of 332 audit tasks to investigate the minimum professional rank and decision aid applicable to perform each task. He also provides a clear view of where the audit judgment is excepted to be correlated with the task .He extended (Felix & Kinney 1982) five audit phase processes (orientation, control structure, test of control, substantive test and audit opinion) by adding a sixth phase (Financial reporting) . each phase identified as follows :
First: Orientation phase
Orientation audit phase can be defined as the process of gathering information about the client's environment. As (AU 311.06) ""The auditor should obtain a level of knowledge of the [issuer's] business that will enable [the auditor] to plan and perform [the] audit in accordance with [PCAOB standards].", the auditor aims to understand the client's geographic, economic and industrial settings, as well as , the auditor should also gain knowledge about the nature and characteristics of the accounting and financial system for the client, and the management integrity and competence. For this phase, Abdolmohammadi (1999) indentified 45 audit tasks that should be achieved during the orientation phase, they are as follows:
Understanding the client's business (12 tasks)
Engagement risk assessment (20 tasks)
Inherent risk (10 tasks)
General considerations (3 tasks)
Abdolmohammadi1(999) finds that the audit judgment is correlated with those task on a different level, he organized the level of judgment under 4 types of judgment. The first is a complete automation decision support system (CA)where very little judgment is needed by the auditor, the second is the decision support system (DSS) that represents an interactive computer software which aims to assess the auditor in making the decision, the third is the knowledge based expert system (KES) where an IF-THEN computer interactive software helps the auditor to make the decision. Finally the human processing (HP) where the auditor is completely relied on his/her judgment to exercise the decision . His results show that 31 out of the 45 tasks inventory are done by senior auditors. Additionally 24 out of 45 of those tasks are unstructured and leave an open space for the audit judgment , 14 tasks were found to be semi-structured and 7 tasks considered as a structured tasks that leave little judgment to be exercised.
Despite the fact that the orientation phase was constructed during the traditional audit (TA) era ,(Ketchcova 2002) adopted his tasks inventory for SSA , she argues that there is no contrary between the two methods during orientation phase which allows to use tasks inventory for SSA method, she also assumed that the same level of personnel are used for both methods, (Bell et al. 2008) consist with (Ketchcova 2002) argument ,as they found a slight increase of the proportion for using high ranked auditors. However, this margin has not significantly increased during the planning phase. To this extend, this dissertation covers the difference between two audit methods during the orientation phase, namely (SSA and TA). A brief summary for the other 5 phases to be discussed below.
Second :The other phases
Abdolmohammadi, (1999) second phase tasks inventory indentifies 41 audit tasks under 12 categories in order to understand the client control structures, a 34 tasks inventory under 6 categories was identified for test of control , 171 tasks inventory for substantive test, where just 1% are unstructured tasks . Another 23 audit tasks were identified for forming an audit opinion. Finally, Abdolmohammadi (1999) identified 12 tasks for financial statements reporting's audit. Table number () in the appendix provides the summary of Abdolmohammadi (1999) results.
Conclusion
This dissertation investigates the decision performance for two audit methods (T.A and SSA) under time pressure. According to (Libby and Lewis 1982) two trends as it falls in the second trend which focuses on the environmental effects on auditor decision making performance from an interaction between two audit methods. Moreover , the scope of this dissertation falls on the first category of (Nelson and Tan 2005), more precisely the (Risk assessments), as the major concern of this dissertation is to compare between "SSA" strategic system audit approach with the traditional audit , in order to assess the client risk during the orientation phase of the audit (Nelson and Tan 2005). In order to do so, Abdolmohammadi, (1999) identified 24 audit tasks that applied during the orientation that relies completely on the auditor's judgment.
Time pressure
"JDM" psychology literature has considerably focused on the effect of time pressure. This attention was also seen in audit literature. As it have examined the progress , consequences and moderators of time pressure from many different aspects . for instants, Coram et al. (2004), investigate the relationship between time budget pressure and the likelihood of auditors engaging in RAQPs., auditors' personality types (Leanne & Gregory 2007) , individuals decisions making under time pressure (Brown and Solomon 1992) , and the impact of time pressure on judgment vs. choice tasks (Smith et al. 1997).
The literature also shows that time pressure related to the task outcome by the following factors:
"JDM" and outcome effectiveness (Andrews and Farris 1972)
"JDM" and outcome efficiency (Smith et al. 1997).
The strategy employed (Libby 1985)
Turnover rate and job satisfaction (Payne et al. 1996)
Interpersonal relationship (Payne et al. 1996)
This dissertation compares the outcome effectiveness of two audit methods under time pressure. Previous studies that investigated the relationship between time pressure and task effectiveness showed conflicted results; the appendix () shows these studies.
The results of those studies found four different types of relationship between time pressure and task effectiveness as follow:
Inverse relationship had been found between time pressure and task effectiveness
No relation found
A positive relationship between time pressure and task effectiveness
An inverted U-shape relationship between time pressure and task effectiveness
Many arguments by researchers aimed to provide the possible explanation for those conflict results. One argument is the failure to capture the same construct. For example, the results of time pressure for experienced participant show a positive relationship with task effectiveness (Andrews and Farris 1972), while different levels of experienced participant show an inverted U-shape relationship with task effectiveness with time pressure. Another argument confirmed the inverted U-shape relationship between time pressure and task effectiveness. However, studies with conflicted results have only captured a limited range of the relationship. A third argument suggests that conflict results may be contingent on other factors in the task environment (Payne et al. 1996) and, thus, moderate the relationship between time pressure and task effectiveness. These factors include:
Variables related to participant personality (personality characteristic)
Variables related to the current situation (task structure )
Variables related to demographic (job experience)
A summery for this study found in table ()
Time pressure and Judgment performance
In psychology, time pressure occurs when the decision maker notes that he/she has a short period of time available to make a decision (Oswald & Kunz 2007). Brown (1999) define time pressure in auditing as when the auditor realized that the time to perform the task is too short and the task completion is an important dimension of the task performance.
As mentioned before, many researches confirm the relationship between time pressure and task effectiveness. Most of these researches used the (Yerkes-Dodson, 1908) laws that suggest an invert U-shape relationship between stress and task effectiveness. As the law suggests that a moderate level of stress can enhance the performance to its optimal level. (McGarath, 1976) argues that two factors are involved to subject task performance to cretin amount of stress, arousal and goal difficulty.
Arousal is defined as the act of stimulating to readiness or to action, while goal difficulty is the amount of effort that the person perform to complete the task. Positive relationship found between arousal and task performance (McGarath, 1976), while a negative relationship found between goal difficulty and task performance (McGarath, 1976, Hirst and Lowy, 1990; Lee et al. 1997, Webb et al. 2010) . This indicates that the invert U-shape relationship between stress and task performance produced by the combined effect of these two factors. However, researchers such as (Scott 1966) argues that low level of arousal factors negatively affect the task performance and vice versa, while a high goal difficulty" such as limited time" even with high arousal can produce a form of stress to participants. Beside, (Brown 1999, McDaniel's 1990) suggest an inverse relationship because for both performance moderators (arousal and goal difficulty) affected under time pressure in deferent direction. Therefore, when combined together, the relation tends to be inversed. This dissertation assumed an inverse relationship supported by (Brown 1999, McDaniel's 1990) findings.
Time pressure and task structure
According to (Yerkes & Dodson 1908) task structure is a moderator factor for the relationship between time pressure (stress) and performance. Task structure has been discussed in details in previous sections. Especially the auditor performance during the absent or the present of task structure (Bonner 1994, Chow et al. 1987, Cushing & Lebbecke, 1986). This dissertation focus on the auditor performance (effectiveness) in the present of two different task structures namely , Top-down and Bottom up task structure under time pressure during orientation phase.
Auditors under unstructured task rely more on identifying the problem, gather the clues, search for alternatives and then issue an audit judgment or choice (Sullivan 1984, Keen & Scott-Morton 1978). Time pressures are expected to negatively influence the audit performance due to heavy reliance on the auditor judgment. Brown (1999), McDaniel's (1990) and McGrath (1976) support this view. On the other hand, auditors under structured task rely more on qualitative analysis of the problem (Sullivan 1984). Therefore, auditor tend to rely more in information processing rather than identifying problems (Bell 1997, O'Donnell 2004). (Brown 1999) conclude from his review to time pressure literature that time pressure has a lesser effect on the auditor performance under structured tasks compared to unstructured ones. However, his finding shows insignificant change in performance between highly structured and semi-structured tasks.
Bottom Up audit structure under Traditional audit
Introduction
As mentioned before, audit firms tend to cut their production cost due to their competitive market .This shift resulted in different strategies by audit firms to adapt with the new competitive environment. Two major changes appeared during the 1980's, the first event was the emergence of audit firms including the big 8 (as was) between each other. Hence, this major change in audit firms' shapes, can be explained as a survival technique in order to cope with the new economic system (Baskerville, 2010). Researchers argued that this merge comes with a many benefits for the mergence. For instance, (Watts and Zimmerman, 1986, Greenwood et al, 1993) argues that the merge appended with the following benefits:
1- Audit firms can reach economic of scale.
2- Help the new formed firm to develop a new brand.
3- Partners are putting all their professional capital together.
4- The easy access to the market from the reputation gained from larger audit firm.
5- More chances to invest in training and new technological process.
6- The ability to serve overseas clients.
The second event was readjusting the audit approach by increasing the usage of analytical procedures (Turley and Cooper, 1991).This shift resulted in more structured audit that aim to increase the control over the audit process. As a result, many researchers investigated the auditors process during analytical procedure in order to execute an audit judgment about the client financial activities . One of the remarkable attempts was provided by (Koonce, 1993). Her work formed the base of the "Bottom Up" structure audit .She introduced a structure of 5 steps to form a cognitive steps for traditional audit that aims to build the decision making process (JDM) during analytical procedure stage for the auditors.
Many researchers followed (Koonce, 1993) model to further investigate about of the 5 steps process.
"JDM" Under Traditional Audit
Traditional audit had relied on a bottom up strategy to test the client's business in order to apply an assessment of the risk that financial statement assertion are materially misstated (Bell 1997, Lemon 2000,Sttaerio 2001).Libby (1985) attempts to introduce a deductive framework to be used during analytical procedures. Libby (1985) borrowed his model from medical field where the purpose is to diagnose patients to give them the right medicine.
He argues that auditors will tempt to generate an assumption about the inter-period change to explain the unexpected fluctuation in the client financial statement. As a result, the auditor based on his/her assumption will relate those fluctuations to their supposed account balances. The auditor then analyses any change in patterns and evaluates the results according to the first assumption. Therefore, if the assumptions about fluctuations failed to confirm the auditor's expectations, then the auditor should refine the previous assumption to a new one that comprises more information and reconciling changes in other accounting measures. Furthermore, the auditor continues to enhance the assumptions until he/she reaches a reliable explanation. Moreover, if the auditor found unsolved problems after this point , the indentified account has to be examined by further testing. (Libby 1985) provide his structure in order to enhance auditors the auditor judgment when assessing the client performance.
(Koonce 1993) suggests that the main factors that derive the analytical procedure under the traditional audit are the diagnostic, sequential and iterative nature. Therefore, Koonce (1993) suggests a model of 5 steps which aims to help the auditor to use decision making process when the results of the analytical procedure are suspicious or unexpected. Moreover, the model target to determine the most potential causes of the unlikely event and to roll out the other causes in order to reach acceptable level of certainty about the audit risk . (Ballou 2005) claims that Koonce's (1993) model had met all the researchers expectation under the traditional audit and that the following researches' where results had fitted with Koonce's (1993) model .Additionally, for Developing an expectation under traditional audit, the auditor builds his/her expectation based on the client financial account changes or any difference appears in the client accounts that materially vary more than the industrial index. The second factor is the rise from a random walk assumption through the client accounts and finally the auditor build his/her expectation on examining the client previous year's end balance. Therefore, the process of developing an expectation under traditional audit and build a mental model is not complicated (Balllou, 2005) , Hence, the auditors used a deductive reasoning for generating their expectation (Koonce 1993). Therefore, when the auditor's expectations are not meeting the reality, the auditor has to generate a different hypothesis that explains the client accounts differences (Libby 1985).
(Koonce 1993) 5 steps model present as follow:
Mental representation
Hypothesis generation
Information searches
Hypothesis evaluation
Decision about sequence actions.
Mental representation
The first step of (Koonce 1993) , about the searches for the factors that influence the formation of the problems, the auditors form a mental representation based on the results of the analytical procedure in order to identify the cause of any fluctuation. The factors that affect this process have been investigated by many researchers. For instant , the level of competence and timing of the source information (Anderson et all. 1994), the auditor experience (Kennedy 1995,Wright & Wright 1997, O'Donnell 2002), riskiness of the audited environment (Anderson et al. 1997), the present of groups or individuals (Bedard and Biggs, 1998), financial and non-financial trends (Cohen et al 2000), the present or absent of accounting measure (Asar et all, 2000), the characteristics of the reviewers during analytical procedure (Ballou 2001), the auditor position (Knapp 2001) the level of fraud risk (NG Green & Simnett, 2001), and The level of the client risk (Mueller and Anderson 2002) .
Hypothesis generation
The second step of (Koonce 1993) model about listing the reasons that may create any unexpected fluctuations of the financial information , the auditor generate a hypothesis that aim to explain the unexpected fluctuations The factors that effect this process has been examined by many researchers. For instant: (Asare & Wright, 1997, Asare et al. 1998, Bedard & Biggs 1998,Church et al, 2001 and Asare & Wright 2003) investigates the generation of the hypothesis process by the auditors. Additionally. (Cohen et al 2000) investigates the hypothesis generation for three accounting fluctuation and (Peecher, 1996) investigates the generation of alternative hypothesis. Other researchers investigates the relation between the number of generated hypotheses with time and subsequent to generate those hypothesises (Bhattacharjee et al, 1999)
Information searches
The third step of (Koonce 1993) model about exploring the information's in order to support or refute the generated or inherited hypothesis/s. Step 2 had described the auditor generated hypothesis that aim to explain the fluctuation, The auditor then look for supporting evidence to his/her hypothesis. The factors that affect the information searches has been examined by many researchers. Such as the application of the analytical procedures (McDanial & Kinney, 1995), the effort invested in information search (O'Donnell 1996), the consumption of budget hours for choosing the test and computing results (Asare & Wright 1997), time spent to perform the information searches (Bhattacharjee et al. 1999), the auditor choices for the audit test(Asare et al.2000),, the presence of fraud in client financial statement (Knapp & Knapp 2001), the difference between provision of balanced evidence verses performing own research (Asare & Wright 2003), and the auditor reliance on the clients explanation or decision aid as a source to explain the auditor hypothesis (Anderson et al. 2003).
Hypothesis evaluation
The fourth step of (Koonce 1993) model about The auditor judgment about the hypothesis under consideration where the auditor select the most likely causes of the fluctuations, this part also had been covered by many researches in order to determine the factors affecting the auditor judgment about his/her hypothesis, for example, presence of inverse base-rate effect (Nelson 1993), The probability estimates for given hypothesis (Anderson et al. 1994,Anderson & Koonce 1995, Asare & Wright 1997, Asare et al. 1998), the initial likelihood judgment about hypothesis (Peecher et al, 1996), the selection of the most likely causes (Asare & Wright 1997, Asare et al. 2000), the assessment of likelihood of material error (Wright & Wright 1997), Time spent to evaluate the hypothesis (Bhattacharjee et al. 1999), allocation of likelihood across hypothesis generated (Charch et al, 2001) and the number of additional evidence to corroborate the hypothesis and number of documentation to explain support for management explanation in working papers (Ballou 2001).
Decision about sequence actions
The fifth and the last stage of (Koonce 1993) model about The determination of consequences for evidence gathering based on the hypothesis evaluation where the auditor choose if further investigation is needed or a reallocation of effort is needed toward specific area in the client account The factors that effect this process has been investigated by many researchers . For instant, the decision about further investigations (McDanial & Kinney, 1995), the determination of planned audit hours by staff level (Wright & Wright 1997), the allocation of audit effort to accounts(Cohen et al. 2000) the change in the extent and time budget hours for the audit test (Glover et al. 2000) and the fraud risk assessment (Knapp & Knapp, 2001)
Conclusion
As mentioned before, (Koonce 1993) model have been investigated deeply by researchers. (Koonce 1993) structure rely on "Bottom -Up" strategy structure where the auditor form his/her explanation to any fluctuations appears during analytical procedure to form a mental representation that aim to build a hypothesis that explain the determine fluctuations, this process followed by the auditor attempted to search for information which support the auditor inherent hypothesis and evaluated it in order to make a decision about the main cause of the fluctuations .However, this approach accused not be interested in how managers manage their business as much as its emphasis on the accounting transaction level to determine the audit risk . Additionally, Bottom up approach also accused to be weak when dealing with ambiguity that associated with unexpected fluctuations from the client (O'Donnell 2005). This ambiguity can be referred to the auditor under Bottom up strategy tend to recognize the symptoms first before he/she generates a hypotheses that is aims to explain those symptoms then gather the information to test the hypothesis . In other words, the auditor start by targeting an account balance and then indentifying the unexpected inter-period changes (fluctuations). The auditor then gather evidences about the nature of the fluctuations and whether it is reasonable or more test is needed. Moreover, to prove that a suspicious account is reasonable for the auditor , auditors must rely on their interpretation of client's business conditions (McDaniel & Kinney, 1995). Where it is hard to tell the difference between unreasonable fluctuations caused by errors or/and misstatement or a fluctuations caused by the normal economic climate changes, which can cause the ambiguity that surround the bottom up audit approach
the analytical procedure process use to be apply during the planning and completion stages under the traditional audit .Traditional audit also had relied on the elements on (SAS No.56) to perform an effective and efficient analytical procedure for the audit client when any unexpected event appears and needs an explanation from the client's management resulted in the auditors directing their attention toward the financial reporting issue at the beginning of the audit process (planning stage) and to help the auditors to identify any issue that could have been missed during the audit as a final stage of the audit (completion stage), as well as, analytical procedure under traditional audit was performed during the substantive test as the only source of the auditing or as a combination with other detailed tests. To this extended, The first three steps of (Koonce 1993) model are related to this dissertation in order to cover the auditor decision making aid for the audit risk factors (Inherent risk and Control risk) under time pressure compared to top down audit strategy ( will covered in later sections) under the same circumstances. Bottom up audit strategy is still the main audit process in Jordanian offices.
