What’S in a Name? An Investigation of the First Stage in the Decision Making Process

ABSTRACT - An extensive review of literature reveals that different terms are used to describe the first stage of the decision-making process. This can be confusing as each term suggests subtle differences in the construct under consideration. This paper investigates the distinctions between these terms. It concludes that a more accurate term for the first stage is 'need activation’. Significantly it was also found that the first stage is a complex sub process within the decision making process.

Citation:

Hasanthi Shalika Subasinghe (1998) ,"What’S in a Name? An Investigation of the First Stage in the Decision Making Process", in AP - Asia Pacific Advances in Consumer Research Volume 3, eds. Kineta Hung and Kent B. Monroe, Provo, UT : Association for Consumer Research, Pages: 204-209.

Asia Pacific Advances in Consumer Research Volume 3, 1998      Pages 204-209

WHAT’S IN A NAME? AN INVESTIGATION OF THE FIRST STAGE IN THE DECISION MAKING PROCESS

Hasanthi Shalika Subasinghe, University of New South Wales, Australia

[The author is greatly indebted to Mary Caldwell, Rob Hall and Paul Henry of University of New South Wales, Australia for providing encouragement.  This paper also benefited from their comments on ealier drafts.]

[Due to space limitations, bibliography is not attached.  Full bibliography is available from author on application.]

ABSTRACT -

An extensive review of literature reveals that different terms are used to describe the first stage of the decision-making process. This can be confusing as each term suggests subtle differences in the construct under consideration. This paper investigates the distinctions between these terms. It concludes that a more accurate term for the first stage is 'need activation’. Significantly it was also found that the first stage is a complex sub process within the decision making process.

INTRODUCTION

Marketers have been interested in the decision making process to understand and to identify leverage points to influence consumer decisions. Decision process models have been popular in the literature as compared to distributive or input-output models. Decision process models can explain 'what goes on inside a consumer’s mind’ whereas the other two approaches failed to provide a detailed explanation of consumer purchase behavior. Distributive models provide information at the aggregate level while input-output models assuming the consumer’s mind is a black box consider only the input and output variables. Both models failed to provide information on an individual’s thought processes when purchasing a product. In contrast, process models provide information on how and why consumers behave the way they do in between the input and output.

Several comprehensive decision making models such as Engel et al.(1968), Howard and Sheth (1969) and Nicosia (1966) have been developed nd popularized since the 1960s. Some of these models have also undergone several revisions to express the complex nature of decision making. For example, Farley & Ring (1970) tested the validity of Howard and Sheth (1969) and then updated it. The basic assumption in these models is that consumers go through a complex process in making purchase decisions.

Explanations of the decision making process are dominated by problem solving literature. Since the beginning of this century, problem solving literature has pointed to a five stage decision making process (Dewey, 1910). These stages include 1) a difficulty is felt, 2) the difficulty is located and defined, 3) possible solutions are suggested, 4) consequences are considered and 5) a solution is considered (see Dewey, 1910 or Dewey, 1933 for more details). Many Consumer Behavior theorists have adopted these five stages in discussing purchase behavior. Katona (1954) suggested that the problem solving behavior can be characterized by the arousal of a problem, deliberation that includes psychological reorganization of the situation, weighting alternatives and finally choose among alternative courses of action (p 31). He used these stages to discuss purchase behavior. Engel et al. (1968) also adapted problem-solving literature in explaining the decision making process. Their five stages of the consumer decision making process are problem recognition, information search, alternative evaluation, purchase and post purchase.

A careful examination of Consumer Behavior literature reveals the use of several terms to discuss the first stage of decision making process. These terms include need recognition (Kerby, 1975, Schiffman & Kanuk, 1978, McDonald, 1994), problem recognition (Engel et al., 1968; Bruner, 1983; Solomon, 1994; Craig Lees et al, 1995 among others), need awareness (Louviere, 1988; Markin, 1974), arousal (Howard, 1977), need arousal (Assael, 1984; Roberts & Lilien, 1992; Rossiter & Percy, 1997), problem arousal (Katona, 1954), activation (Kindra et al, 1994) problem identification (Brim et al, 1962) and problem perception (Williams, 1982). The variation of terminology in the literature is confusing to many. A deeper understanding of each term is required to consider its use or misuse and its appropriateness. An understanding of the differences between these terms is crucial for the advancement of the decision-making literature.

This paper focuses on the first stage of the decision making process. The main focus of this paper is to address the following areas: the research imperative, the first stage in the decision-making process, the most appropriate term to denote the first stage of the decision making process. Finally, a new model of the first stage of decision making process is proposed.

1. The research imperative

This first stage of decision making research is crucial for several reasons. Firstly, consumers need to identify that they have a need to purchase a product. This need can be defined as the product category need. Marketers strive to satisfy these consumer needs. Some consumers tend to go through different decision stages sequentially or simultaneously in purchasing products. Activation of purchase need is required for impulse purchases. These needs can be unconscious, preconscious or conscious and it takes place within few seconds.

Secondly, the remaining of the decision stages depends on the activation of a need. Identification and activation of needs increase the likelihood that consumers will engage in information search, evaluate alternatives and eventually purchase the product. Consumers do not have to go through each stage sequentially. However, the first step can not be skipped for any reason. Mowen’s (1988) three perspectives of decision making also require need activation for a purchase to occur.

Thirdly, understanding can help marketers to reduce any delays that can occur at this stage. Greenleaf & Lehman (1995) found that fifty-three percent of the delay in the decision makng process occurred at this stage. Although marketers have relatively little control over some delay factors, they can develop strategies or facilitate information to overcome these constraints.

Fourthly, the identification and understanding of the first stage of the decision-making process can help marketers design marketing mix elements that help trigger the need for a product. Effective use of marketing strategies to influence the need activation can result in an instant purchase for impulse purchased products or the activation of the rest of decision process for other products.

Finally, this stage is under researched (for example, Roberts & Lilien, 1992) or the specific details are often not discussed in detail (for example, O’Brien, 1971). Furthermore, some researchers have not included it in their discussion at all (for example, Udell, 1966). Although most researchers have investigated different stages in the decision process, there is very little research to understand this crucial first stage.

2. The first stage of the decision-making process

As mentioned already, discussion of the first stage of the decision making process use several terms, namely problem recognition, need recognition, arousal, need arousal, problem arousal, need awareness, activation, problem identification and problem perception. The definitions of these terms are discussed in the following section.

Problem recognition:

This is the most widely used term in the literature. Engel et al (1968) and Sirgy (1983) have formulated the two main definitions of problem recognition.

a) Engel, Blackwell & Kollat: The most common definition of problem recognition suggests that "it results when a consumer recognizes a difference of sufficient magnitude between what is perceived as the desired state of affairs and what is perceived as the actual state of affairs" (Engel et al., 1968: 361). By comparing these actual and desired states, a consumer tries to balance these two states. Bruner (1986) used homeostasis theory to explain the balancing nature of actual and desired states in problem recognition. This balancing task is activated by the degree of discrepancy between these two states and the importance of the identification of the discrepancy. However, the sufficient magnitude of difference required is not yet defined in the literature. Furthermore, this definition is individual in nature as it provides less emphasis on the sociological factors. It was suggested that one or more motives underlie the desired state. Actual state is also said to involve many variables interacting in complex ways.

Engel et al (1968) suggest that problem recognition involves two outcomes. Firstly, the process halts if consumers do not perceive a difference in sufficient magnitude between the actual state and desired state. The process also stops when consumers perceive need fulfillment is strongly hindered by constraints such as finance, time, energy etc. Secondly, problem is perceived and proceeds with the process depending on the type of decision making. The nature of the decision making process ranges from a relatively simple response to one that is more complicated. For instance, a simple and standard learning process can arise if a product is habitually purchased. In contrast, a more complex type of decision and problem recognition can occur when the consumer is highly involved with the product.

The problem recognition state is nique due to several aspects. Firstly, problem recognition processes may simultaneously involve many variables i.e. attitudes, motivation, values etc. Thus, the problem recognition process is more complex than motivation. Secondly, the problem recognition may involve complex weightings of all these variables. Finally, the problem recognition process is more than simply awareness and interest as suggested by the diffusion literature. Most importantly, problem recognition goes beyond awareness of a product as it triggers the rest of the decision process.

b)  Sirgy : The second definition was put forward by Sirgy (1983, 1985, 1987) using cognitive psychology literature. According to him, problem recognition is a function of the discrepancy between the valence level of perceived performance of ones current product and the valence level of a referent. Perceived performance of the current product (perceived value= PV) and the referent (evoked value= EV) provides congruent or incongruent states. Congruity theory was used to explain the differences between these states and their relation to problem recognition. Congruity was defined as the degree of discrepancy between the valence of the perceived objet (PV) and the valence of the referent state (EV). Valence of the perceived object is denoted by the desirability weight of the evaluated product. The weight contributes to the desirability of the product. Perceived value provides the evaluation of the current product and evoked value provides the evaluations of the ideal product.

Sirgy (1983) discussed the perceptual congruity and evaluative congruity. Information processing directed to evaluate the stimulus object is defined as evaluative congruity and information processing directed to interpret stimulus object is defined as perceived congruity. Sirgy (1987) mainly concentrated on evaluative congruency and identified four evaluative congruity conditions. These are a) positive incongruity- significant discrepancy between a positively perceived product performance and negative evoked value, b) positive congruity- significant discrepancy between a positively perceived product performance and positive evoked value, c) negative congruity- significant discrepancy between a negatively valued perceived product performance and negative evoked value, d) negative incongruity- significant discrepancy between a negatively perceived product performance and positively valued referent state. He also found evidence for varying degrees of problem recognition depending on the evaluative congruity condition (Figure 1).

Furthermore, Sirgy (1983) posits that problem recognition can be categorized as need recognition and opportunity recognition. Need recognition occurs when consumers evaluate its current product whereas in opportunity recognition consumers evaluate a new product or unused product. Therefore, Sirgy’s notion of problem recognition is different from Engel et al (1968).

Need recognition:

McDonald (1994) defines need recognition as the difference between the actual state and the desired state. Schiffman & Kanuk (1978) define need recognition as the recognition or the realization of a need by the consumer that there is a difference between 'what is’ and 'what should be.’ The term 'what is’ is similar to a consumer’s current state or the actual state and 'what should be’ is similar to desired state. Therefore, these two definitions are similar to problem recognition as defined by Engel et al. (1968). However, the only difference is the use of a different term to define the first stage. However, Sirgy (1983) suggests that need recognition occurs when consumers evaluate a current product rather than a new or unused product. Bruner & Pomazal (1988) adopted Sirgy’s (1983) definition and defined need recognition as the problem recognition due to the changes in the actual state rather than the desired state. Thus, it is clear that there is a discrepancy in agreeing on the term need recognition in the literature.

Need awareness:

Louviere (1988 ) does not clearly define need awareness. Although this stage is included in the decision process, he failed to detail the definition or dynamics of this stage. Awareness is the first stage of Roger’s (1962) model in the diffusion literature. Need awareness is not clearly defined in the literature. However, it suggests the awareness of one’s needs which leads to the rest of the decision making process.

FIGURE 1

Arousal:

Howard (1977) defines arousal as the first stage and it encompasses the need activation. Engel et al. (1968) also discussed arousal in the context of motivation. They posited arousal and perception to be closely related. Arousal occurs because of drive activation, environmental stimulation and autistic thinking. However, most of these arousal states are not easily accessible.

Need arousal:

Assael (1981) discusses need arousal as a psychological set toward the prospective purchase. Roberts & Lilian (1993) discuss the need that is activated or aroused through internal or external stimuli. Assael (1981) considers one’s needs relative to the product category need and ones’ attitude toward various brands. However, arousal of a need is most often associated with the category need and not the attitude toward various brands. Habit and familiarity with a product category can lead consumers to refer to brands. However, even on this occasion a consumer unconsciously considers the category needs when discussing brand needs.

Problem arousal:

Katona (1954) does not clearly define problem arousal. He assumed that consumers go through a problem solving process. Purchase is initiated by problem arousal.

Activation:

Kindra et al (1994) used activation as the first stage of the decision process. They posited that once a consumer perceives that currently consumed goods cannot adequately fulfill a need or solve a problem, a gap arises between the desired state of affairs and the current situation. This gap creates tension. Then, there is a tendency to alleviate this tension. When the consumer decides to regain the desired state, a person has activated the decision process that leads to consumption and problem solution. This seems to draw from Engel et al.’s (1968) definition of problem recognition and motivation theory of drive reduction in psychology.

Problem identification:

Problem identification is defined as the person’s recognition that he is confronted with a problematic situation that requires attention and definition (Brim et al, 1962). Problem identification relies on consumer’s identification of a problem rather than a need to purchase a product. They also posit that there are individual differences across different problems and time spent on each problem.

Problem perception:

Problem perception is another term used in discussing the first stage. William (1982) states that problem perception is what alerts us that we require something. He posits that the perception of a problem comes to us through our senses and involves the recognition that things are not exactly as we would like them to be (p 27, 29). These consumers tend to perceive that they have a problem rather than a need in making purchases. The purchases are made to solve the perceived problem.

Table of definitions

Table 1 illustrates the preceding definitions and terms.

It is vital to consider all the terms used in the literature to denote the first stage of the decision-making process. Different terms in the literature bear a resemblance to its origin of the theory. The term containing 'problem’ originated in the problem solving literature and assumes that consumers go through a problem solving activity while 'need’ originated in psychology.

Critical evaluation of the above shows that they can be classified as dealing with arousal, recognition and identification, activation, awaeness and perception of needs or problem. The summary of these terms clearly demonstrates the confusion in the literature with regard to the first stage of the decision-making process. The confusion in the literature can be overcome by having a single term to denote the first stage of the decision making process.

3. Most appropriate term for the first stage of the decision making process

The summary table in the previous section demonstrates the number of terms used and their origins. These terms do not adequately capture the dynamics of the first stage of the decision making process. Thus, it is very important to consider a term that can capture more adequately the dynamics of the first stage in the decision making process. A careful examination of the terms discussed in the previous section could provide some guidance. The following sections examine the origin of the first stage of the decision making process. The origins clearly lead us to analyse the terms, 'problem’ and 'need.’ Following this, the nature of arousal, awareness, recognition, identification and activation are examined. This could provide suggestions as to the extent to which each can be used to represent the first stage of the decision making process.

3.1 The terms 'problem’ or 'need’

Decision making literature includes the specification of problem or needs as part of the first stage. The terms 'problem’ and 'needs’ are discussed below. Analysis of these terms is a vital component in considering the what term is more appropriate.

Problem:

The word 'problem’ assumes that consumers have a difficulty associated with purchases. The term relates to, anything, (matter, person etc.) that is difficult to deal with solve or overcome (Collins Dictionary). Some tend to think of a problem with the current product as the only factor in influencing this stage. Rossiter and Percy (1997) named need arousal as the first stage and argued that they rejected problem recognition as it is not always triggered by a "problem" with the current product (p 100). However, Engel et al. (1968) stated that it is triggered not only by a problem with the current product but also due to other factors. Some of the factors include new needs, dissatisfaction with the current product, reference group influence, changes in the financial status etc. There are others who found that this stage is triggered not only because of the current product problem. Punj and Srinivasan (1992) found new needs and problems with the current product influencing problem recognition. Wilkie and Dickson (1979) found appliance failure, replacement of current unit (due to repair, needed a new one) and residential move influencing the white goods purchase needs. Perhaps it is the use of the term 'problem’ that may be confusing to many. Furthermore, the term 'problem’ carries negative connotations that are inappropriate to describe the first state as it includes negative as well as positive connotations. This can be avoided by using the term 'need.’

TABLE 1

Needs:

Need is defined in different ways. Need is often used to denote a drive or some inner state that initiates a drive. The nature of needs provides a better understanding of the meaning of needs.

The nature of needs:

* Innate and learned needsBThe literature reveals two different views of needs. First, needs are considered innate and these needs are more biological and physiological in nature. An individual is born with needs. Second, needs are considered learned. These needs are learned through the socialization process. McCelled’s theory of motivation suggests that all needs are learned. These needs are mostly acquired through association that is explained by learning theorists as conditioning. This conditioning occurs through continuity and reinforcement. The basic learned needs are the need for achievement, affiliation and power. Those with high achievement needs strive for success and take responsibility for solving problems. The need for affiliation makes a person associate with others and is similar to Maslow’s belongings need. The power need makes one obtain and exercise control over others. McClelland et al’s (1953) affective arousal model suggests that cues paired with adoption level discrepancies innately produce affect. These affective states are positive or negative in nature. Murray’s motivational system assumes primary needs are biological in nature and secondary needs are learned.

* Latent and manifested needsBNeeds can be manifested or latent. Manifested needs are known to consumers and are part of unmet needs. Manifested needs are mostly primary needs and consumers are conscious of these needs. Latent needs are not known consciously. These latent needs are most often identified through in-depth or indirect probing. Latent needs are also stored in the preconscious memory.

* Conscious and unconscious needs- An individual has an immediate retrospective awareness of conscious needs. Individuals can recall these needs and also can verbalize these needs. All other needs that can not be verbalized are considered as unconscious needs. Murray (1938) suggests that unconscious needs are expressed in dreams, visions, emotional outbursts and unpremeditated acts, slips of the tongue and pen, absent-minded gestures, projections (illusions, delusions and beliefs.

* Primary and secondary needsBMurray’s motivational system assumes primary needs are biological in nature and secondary needs are learned. Primary needs involve three kinds of activities- the intake of substances, the output of substances and withdrawal behavior. Each secondary need is learned and we have different experiences therefore individual differences exist in secondary needs. McClelland et al (1953) disagrees with this distinction and suggest that all needs are learned and the difference can be shown by discussing the cues involved. They suggest that cues involved in primary needs are mainly biological and secondary needs are mainly social in nature.

* Positive and negative needs- Murray (1938) divides positive needs into adient needs and contrient needs. Adient needs cause an individual to approach a liked object and contrient needs cause an individual to approach a disliked object in order to dominate aggressively, abuse, injure or destroy it. McClelland et al (1953) also discuss the positive and negative nature of affect in their affective arousal model. They suggest that positive affect is the result of smaller discrepancies of a sensory or perceptual event from the adaptation level of organism. Negative affect is the result of larger discrepancies. For example, sweetness is a positive affect and bitterness is a negative affect. Positive affect results in approach motives while negative affect results in avoidance motives.

* Hierarchy of needs BMaslow (1943) suggests that humans satisfy needs in a hierarchical fashion and that needs consist of physiological needs, safety needs, love and belonginess needs, esteem needs, self actualization needs. Physiological needs consist of the basic needs such as food, water, and shelter. Safety needs include seeking physical safety and security. Love need includes the desire to be loved, to belong to a group. Esteem need includes the ones desire for status and superiority. Self-actualization need involves the desire for self-fulfillment. Maslow suggests that needs are innate and people try to satisfy lower order needs before the higher order needs. Most powerful needs dominate the consciousness while the less important needs are denied or forgotten. McClelland et al (1953) also agrees with the notion of hierarchy. They suggest that 'motivatin does seem to activate particular response hierarchies in the organism, even though this does not always in increase in overt activity’ (p 90).

* Deficiency and growth needs- Maslow (1968) discussed deficiency motives and growth motives. Deficiency motives seek to reduce tension and seek equilibrium. These deficiency motives are short term in nature and homeostasis was discussed as a tension reducing strategy. Deficiency need includes physiological, security and esteem needs. Growth needs are long term and continual in nature. Maslow’s self-actualization need is especially considered as a growth need. Growth motives maintain tension in the interest of distant and often unattainable goals.

* Stimulation levels existBStimulation is vital for the needs to be triggered and there are different levels of stimulation levels.

In summary, it is very evident that need is a complex construct. These explanations of the nature of needs clearly show that the term 'need’ does explain more than the term 'problem.’ The use of the term problem may not be able to capture all the dynamics of the first stage of the decision making process.

3.2 Is a single term more appropriate?

Next, it is imperative to analyze the terms such as arousal, awareness, recognition, identification and activation to understand which construct can explain the dynamics of the first stage.

* Arousal: Arousal refers to the physical excitement and the physiological changes that are part of emotion in psychology (Bootzin et al, 1991). Collins Dictionary defines arousal as evokes or elicits a reaction, emotion or response. Arousal can be physiological and or psychological. Arousal is based on the arousal theory in motivation. However, arousal does not involve much deliberation. The arousal state would influence the activation of the rest of the decision process. Urgency of the aroused need also influences the consumer’s reaction to the need state. Therefore, arousal is not sufficient to engage in purchase for impulse purchases or to go through the rest of the decision process. However, the dictionary definition encompasses the elicitation of reaction meaning that it is capable of activating the rest of the decision stages.

* Awareness: Awareness can be defined as having knowledge (Collins Dictionary). Needs must be aroused for the consumer to become aware of the needs. It can be as a result of physiological arousal or psychological arousal; the consumer has to use some cognitive effort to be aware of the need. Once the need is aroused, it may need to reach a certain level for the consumer to be aware of the need. Therefore, arousal of a need must precede the awareness of the need.

* Recognition: Recognition is the act of recognizing or remembering (Collins Dictionary). A consumer’s recognition of a need or a problem requires more cognition. However, for the recognition to occur a consumer’s need must be aroused. Awareness also helps a consumer to recognize the need. Therefore, arousal and awareness must precede the recognition of the need.

* Identification: Identification refers to the act of identifying or the state of being identified (Collins Dictionary). The psychology literature defines identification as the process of recognizing specific objects as the result of remembering (Bootzin et al, 1991). A cnsumer must recognize the need for the identification to occur. Need identification can direct the consumer what product category to consider for a purchase. A consumer may not take action even after the identification of the need. There must be some energy to elicit action. Therefore, identification is not sufficient without eliciting action for a marketer.

* Activation: Activation is defined as being active or capable of being active (Collins Dictionary). Krober-Riel (1971) posits activation as a means of manipulation and a consumer cannot escape from it. Marketers are interested in activating needs of consumers. Activation of needs makes the consumer go though the rest of the decision process. This activation can reduce the delay that has occurred at the first stage. Identification of the product category need is necessary for the need to be activated.

The above discussion provides clear evidence that 'need’ is a better term than 'problem’ as consumers tend to fulfill needs. Further, under impulse purchasing situations, consumers try to satisfy their urge rather than solve a problem. Impulse purchase situations also require need activation to occur. However, under certain circumstances, consumers tend to skip most other stages of the decision process. Moreover, the above discussion demonstrates that one term alone can not capture the dynamics of the process at the first stage. Need activation seems to comprise all the elements and it is the best term out of all other terms. However, even this one term can not explain the actual process that takes place, as there is a process within a process.

The new model of the first stage in the decision process

The first stage of the decision making process is important for marketers only if the momentum of the needs is increased to get the consumers to go through the rest of the decision making process. A new model which includes the increase in need momentum is proposed in the next section. These different stages in the first stage of the decision making process has not been captured in the literature previously. The new model encompasses the different thought processes a consumer goes through before going through the rest of the decision process. These steps within the need activation process capture the necessary requirements for the activation to occur. These activation process steps were derived by following analyzing of the terms in the previous sections.

The activation process consists of following steps:

Arousal => Awareness => Recognition => Identification => Activation

This new process is called the need activation process. Although Kindra et al (1994) use the term 'activation process,’ they do not provide an explanation of the activation process. Roberts and Lillian (1993), in discussing need arousal, use the term 'need activation (arousal)’ encompassing the activation of needs. However, they do not provide the relationship between arousal and activation. Kroeber-Riel (1971) posits that activation is a strong arousal state involving cognition. The term need activation captures what the marketer is finally interested in at this first stage. This need activation process explores the possibility of a process within the first stage of the process and it provides more insights into the first stage of the decision making process. However, all the elements of the process need to be explored to have a good understanding.Thus, the further study of need activation process is required.

CONCLUSION

This paper has critically evaluated different terms used in the literature to describe the crucial first stage in the decision process. Several terms are critically evaluated to consider the use and misuse and the appropriateness of the terms. It is proposed that we can no longer consider the first stage using one single term as this stage includes a process. This process can be best described as the need activation process. As marketers are more interested in the activation rather than arousal, awareness, identification or recognition, activation is a much better term and captures the psychological and physiological process. Furthermore, 'need’ is preferred over the term 'problem’ as marketers are interested in needs rather than problems. Therefore, this paper contributes to the literature by highlighting the distinctions and exploring the need activation process.

The understanding of this need activation process provides a marketer the necessary information to create strategies to activate needs. Further, need activation can reduce the delay that is reported in the literature. For example, the delay factors such as financial constraints and time constraints can be overcome by introducing easy payment options, providing options to do shopping from home etc.

There are several limitations in this paper. Firstly, it can not do justice to the complex nature of the first stage. Secondly, we can not conclude that this proposed model of the need activation process is the best without empirical evidence. Nevertheless, a comprehensive analysis of all the various accounts of the need activation process can provide a rich source of information for a marketer.

----------------------------------------