Requirements of Grade B
Literature Review
"Like death and taxes, organizational conflict will always be with us" (Putnam and Poole, 1987). Conflict is ever-present in organizations all over the world. Because of its ominous presence, conflict plays a very significant role in an organization. According to Harris (1993) this is due to three reasons. The 1990s represents a very turbulent era, increasing the occurrence of conflict throughout organizations. Secondly, conflict cuts across every level of organization, no one is left unaffected. Thirdly, organizations are conglomerates of individuals, groups, departments, special interest and divisions.

From a communication perspective, conflict as expressed by Hocker and Wilmot (1985), is "an expressed struggle between at least two interdependent parties who perceive incompatible goals, scarce rewards and interferences from the other party in achieving their goals." Keeping in mind that organizations consist of people of differing interests and personalities, clashes among these people are inevitable. Thus, although conflict can be beneficial for the company, it has to be kept under control so that it stays at a manageable level and does not turn into dysfunctional conflict.
The most popular approach of managing conflict styles comes in the form of a Managerial Grid formulated by Blake and Mouton in 1964. Conflict style refers to a characteristic mode or habitual way that a person handles a dispute. Some research in this area, however, also treats style either as an orientation towards conflicts, categories of conflict strategies and tactics or planned strategies aimed at attaining a particular goal (Folger & Poole, 1984).

The grid offers a five-category scheme that is based on the two dimensions of concern for self and concern for others. This scheme presented the first conceptual scheme for classifying the modes or styles of handling interpersonal conflict into five categories: problem-solving, smoothing, forcing, withdrawing and sharing.

The two dimensions of concern for self and concern for others form the axes of the grid. The former (occupying the y-axis) explains the degree (high or low) to which a person attempts to satisfy his or her own concern. The latter (occupying the x-axis) explains the degree to which a person wants to clarify the concerns of others (Rahim and Bonoma, 1979). Similarly, the y-axis also represents the degree of assertiveness and the x-axis, the degree of cooperativesness of the person handling the conflict (Hutchinson, 1992).

It is important to note that although Blake and Mouton (1964) treated the five styles as interrelated points along the two above mentioned continua, Putnam (1988), pointed out that researchers like Vliert and Prein view the five styles as distinct categories for classifying behaviors and ignore or minimize the two dimensions. Other authors conceive of the five styles as interrelated choices that grow out of each person's cognitive emphasis on the two dimensions.

The grid was later refined by others like Filley (1975), Hall (1986), and Killman & Thomas (1977). Using the grid as a point of reference, they developed several variations of the five category scheme. Although they named the five categories differently, they essentially represent the same styles of conflict management developed by Blake and Mouton in 1964. One such variation groups people according to whether they are more likely to compete by using position power, verbal dominance and assertiveness in a conflict situation, compromise by being a mediator to split the difference, accommodate by trying to smooth over and trivialize the conflict, avoid by removing oneself physically from the conflict scene or collaborate by being a problem solver and working for an integrative solution (Harris 1993).

Similarly, Neher (1997) categorized the five conflict styles into avoiding, which is low on both concerns (self and others)., smoothing or accommodating, that is low on concern for oneself but high on concern for others, compromising, that is moderate on both concerns, forcing, that is high on concern for oneself but low on concern for others and lastly, confronting, that is high on both concerns.

The five conflict styles, can be cast in terms of a win-lose dichotomy (Neher, 1997). Each of the five styles described above can be assessed in terms of whether the parties involved in the conflict situation "win or lose". When one person wins in a conflict, another must lose. Using the style of avoiding and compromising would represent a lose-lose situation while accommodating and forcing represent a lose-win situation. Confronting represents a win-win solution and this is the only truly integrative style that does not present conflict as a zero-sum game.

In 1992, Putnam and Wilson came up with an instrument called the Organizational Communication Conflict Instrument (OCCI) to measure conflict strategies. The scale asked respondents to respond on a 7-item Likert type scale, how they would act in a particular conflict situation. The results saw a revision of the five categories to three categories: nonconfrontation strategies, (which included avoiding and smoothing), solution-oriented strategies (which included confrontation and compromise) and control strategies (which included forcing).

The OCCI presented a more contemporary approach to measuring conflict resolution by shifting the focus from an interpersonal focus as done by Blake and Mouton (1964), Hall (1969), Lawrence and Lorsh (1967), and Kilmann (1974) to an organizational focus where the questions were framed to reflect an organizational setting. However, because the instrument sought to conceptualize resolution in terms of the strategies of the participants with respect to specific situations only, it fell prey to two shortcomings. Firstly, it did not take into account the role of individual differences in influencing the choice of conflict styles. Secondly, it did not take into account the fact that different situations would produce different results as to the conflict style that someone would choose (King and Miles, 1990).

Another more recent scale, the Rahim Organizational Conflict Inventory, sought to measure the style configuration of people in organizations. The difference between the two scales was that the latter was designed to measure a single characteristic style that "may reflect a predisposition, habit, or stable, internal preference" for a certain style (Wilson & Waltman, 1988). Thus, this scale would be more aptly measure the style preference of an individual because it takes into account not just one solitary incident, but looks across all conflict incidents.

The problems of these two scales led King and Miles (1990) to conduct research to find out whether people approached conflict situations systematically or situationally. Situational contingencies, for example, might determine the way or influence the way a person selects a particular strategy. Neher (1997) pointed out that factors like structure of an organization, expectations about the relationship between conflicting individuals, views held on reciprocity, the presence of shared goals, the setting of awareness, and the presence of audiences, might affect the way an individual chooses a conflict style.

The contingency approach, the result of situational differences, also reflects the fact that people have patterns that they follow during conflicts. The patterns reflect what they feel make sense to them at the time of the conflict (Hocker & Wilmot, 1985). In most organizations for example, collaboration and cooperation are required to achieve certain goals (Harris, 1993). However, there are times when conflicts cannot be solved the usual way. In this case, parties who are aware of the five styles may choose one which they feel best fits the situation. Evidence of support for the contingency theory came in the from of the rejection of one best style for all conflict situations (Putnam 1988).

The results of the research done by King and Miles (1990) however were unable to clearly answer whether styles and strategies were truly relatively stable or situationally dependent. They did however conclude that either context played an integral part in conflict management and that the choice of conflict styles are indeed stable, or that if conflict styles and strategies are indeed stable, the current instruments are unable to capture that stability.

The limitations of the current instruments can be attributed to several factors. Much of the research has incurred problems in the conceptualization of conflict style. For example, a particular conflict style like forcing can be communicated in a variety of ways. Individuals can argue persistently, raise their voice, use a wide range of words, nonverbal cues and conversation patterns. Thus, the way in which an individual forces a decision may be more significant than the use of forcing as a strategy or dominant style (Putnam & Poole, 1987)

Another problem stems from the lack of attention to the way a conflict style may shift due to interaction among people. A dominant orientation to conflict usually alters once interaction progresses. Thus, although instruments like the OCCI measure conflict styles of a particular context, it does not measure how a conflict style could change within that particular context itself.

Putnam and Poole (1987) also noted that communication was often reduced to misunderstandings or ambiguities, thus oversimplifying the role of interaction in conflict situation. In much of the past research done, the choice of conflict style centered exclusively on the isolated cognitive systems of each disputant. Researcher felt that studies that incorporate inter-group and inter-organizational levels of analysis would give a greater insight into the choice of conflict style.

For our study in particular, we are interested in the differences in conflict-handling styles between men and women. Much of the research uncovered points to many interesting differences in personality, cognitive and behavioral gender differences. Both research and everyday experiences indicate that there are differences between males and females. Whether these differences are based on biology, social learning or a combination of the two remains elusive.

Bem did a study in 1974 to identify the content of sex-role stereo types. She then studied the effects of individual differences in adherence to these sex-role behaviors. It was found that men were believed to possess more leadership ability and to be more assertive, dominant, and forceful than women. On the other hand, women were believed to be more gentle, tender, compassionate, and warmer than men.
Social psychologists have also found that children between the ages of four and seven have already acquired the principles of gender consistency. At such a young age, they are already able to grasp the concept that gender is a basic attribute of an individual that remains the same over time and that it is unaffected by external factors. Understanding gender identities depend very much on socialization experiences in childhood. By the adolescence period, behavior that can be predicted by gender will already be well developed and practiced. This is the time when gender is firmly lodged as part of a person's social identity. Thus, behavior will then be consistent with gender roles.

It is not surprising then that when people reach adulthood and step into their occupational roles, they bring along with them gender-based behaviors. The job market becomes a place where gender roles are greatly manifested. To some extent, societal expectations about appropriate gender-based behaviors have influenced men and women to respond differently in the work context.

One area where men and women behave differently is when interacting with others. Bountiful research has found that men and women have vastly different styles in interpersonal communication. For instance, women are more likely than men to share rewards (Major & Deaux, 1982) or to deprive themselves in order to help someone else (Leventhal & Anderson, 1970).

Another major area of concern in the workplace is the differences in managerial styles of men and women. Women's leadership styles are characterized by connective and interactive behavior. In other words, women prefer collaborating, consultating, and negotiating with others while men value competition, individual achievement and demands (Blumen, 1988; Rosener, 1990).
In a study done by Rozier and Hersch-Cochran in 1996 on gender differences in managerial characteristics in a female-dominated health profession, it was found that even in the female-dominated industry, men still tend to use masculine characteristics more than the women.

Other research in this area attributes the differences to self-esteem. For men, self-esteem is linked to individual achievements, but for women, self-esteem is linked more to their interpersonal achievements with specific others (Josephs, Markus & Tafarodi, 1992). According to Margalit and Eysenck (1990), women tend to outperform men in social skills.

The research done by King and Miles (1990) was also aimed at an exploratory investigation of individual differences along the lines of self-esteem, Machiavellianism and the Protestant Ethic. Machiavellianism refers to gamesmanship and manipulation to gain an advantage over an another. Protestant Ethic is a personality variable that focuses on the degree to which a person feels that personal worth results from self-sacrificing work or occupational achievements. They were interested to see how individual differences affected conflict management choices. Of the three, only self-esteem displayed significant results in determining the type of conflict management style would be undertaken - respondents with low self-esteem tended to adopt avoiding behavior and were less likely to engage in dominating, compromising, and integrating behaviors.

There is also much research that indicates that the inherent differences in the personality of men and women that lead to many of the above-mentioned behavioral differences. Fesingold (1992) came up with four models that contrasted biological and environmental factors in their shaping of sex differences in personality and behavior. The biological model posits that gender differences are innate. The sociocultural model proposes that social and cultural factors exert pressure on men and women to behave differently. The expectancy model emphasizes that cultural influences on perceptions about different sexes produce different expectancies. The artifact model suggests that sexual stereotypes influence what we believe about ourselves.

Our study focuses on the conflict management aspect of interpersonal communication within an organization. According to Mintzberg (1973), conflict resolution is one of the concerns of managers. However, our study is more interested in conflict resolution at all levels of the hierarchy within an organization rather than focusing only on the managers. This is because conflict exists within formal and informal modes of communication.

Given the abundance of literature on gender-based behaviors, it would be interesting to study how men and women differ in their conflict resolution styles. There have been contradictory findings on sex differences as predictors of conflict management style.

From our literature search, there appears to be two main predictors of conflict management style preference, namely gender and psychological type. A research conducted by Mills, Robey and Smith (1985) replicated previous research that had found that conflict styles are rooted in deeper human constructs like personality or values. They aimed to establish a relationship between conflict resolution style and Jungian personality dimensions. Jung's approach to the study of conflict handling was to describe the process by which people perceive information and make judgements. The typology offers two forms of perception -- sensing and feeling, and two forms of judgement -- thinking and feeling. According to Jung, one's attitude can also be extraverted or introverted. These dimensions make up one's psychological type.

Findings from the study done by Mills, et al, were consistent with previous findings. There were significant correlations between thinking and assertiveness, between feeling and cooperativeness, and between thinking and distribution. Feeling was also positively correlated with the accommodating mode of conflict management. Extraversion was positively correlated to assertiveness while introversion was positively correlated with the avoidance mode of conflict management.

Although we can very well conclude from all these research that psychological type is indeed an accurate predictor of conflict-handling style, it must also be noted that gender is also a very accurate predictor of psychological type. In addition to the above findings that link psychological type to conflict handling, Mills, Robey, & Smith (1985) also found that males were predominantly "thinkers" and females were predominantly "feelers". "Feeling" is to attach subjective, personal value to an object or phenomenon whereas "thinking" is characterized by analytical logic and reasoning. Their research suggests that in conflict situations, women tend to be cooperative or accommodative while men tend to be competitive.

Women tend to adopt cooperative strategies and tactics which allow the other side to save face, and often, they also address the legitimacy of their concerns. Men, on the other hand, are more coercive and forcing (Dowd, 1991). According to Gayle (1991), this difference may be rooted in traditional power and status differences.

Similarly, Todd-Mancillas and Rossi (1985) found a significant difference in the communication strategies of male and female managers. Male managers were more likely than female managers to use power and authority to resolve conflict with their subordinates whereas female managers resort to these only when interaction and communication failed.
Papa and Natalle (1989) compared interpersonal communication between male-male, female-female, and male-female dyads, and found that communication were consistent with gender stereotypes. Female-female dyads focus on bargaining to achieve resolution. Conversely, male-male dyads refused to resort to bargaining to resolve conflicts.

Although much research posits that gender is very much an indicator of conflict-handling behavior, other research seems to suggest that the relationship is not so clear. A research by Duane (1989) found that men and women did not differ on their use of collaborative or compromising modes of conflict management. Paula S. Sorenson, Katherine Hawkins and Ritch L. Sorenson (1995) found that while psychological type may be a more powerful indicator of conflict style preference than gender, neither psychological type nor gender accounted for a substantial amount of variance in conflict style preference.

One research done by Mc Kenna and Richardson in 1995 outlines the conflict-handling style of a representative sample of 303 Singaporeans. Besides using gender as a variable, the study also tested other variables such as age, role status and occupational differences using the Thomas-Kilmann conflict mode instrument. This research was found to be particularly relevant because of its local context. Predominant styles for both men and women were found to be compromising and avoiding respectively. Men tended to adopt compromising strategies while women tended to use avoiding strategies.

Research done on this topic presents contradictory findings. Therefore, in our study, we hypothesize that there is a difference between men and women's conflict-handling styles in an organizational setting. In order to clarify what are the most common styles of conflict management adopted by men and women in an organizational setting in Singapore, we also pose a research question.

Hypothesis And Research Question

Hypothesis:
Men and women adopt different conflict resolution styles in an organizational setting.

Research Questions:
(i) What is the most common conflict resolution style adopted by men in an organization?
(ii) What is the most common conflict resolution style adopted by women in an organization?

Requirements of Grade A
Method
Subject
As the main purpose of our study is to establish the effects of gender on the type of conflict resolution preferred by people in an organizational setting, the subject of our study is any employee working in an organization with a hierarchical structure.

The Independent and Dependent Variables
The independent variable in our study is gender (male or female).

The 30-item Organizational Communication Conflict Instrument (OCCI) designed by Putnam and Wilson was to operationalize our dependent variables, the different conflict strategies. This instrument conceptualizes conflict resolution in terms of strategies and the instrument measures three conflict strategies. Nonconfrontation strategies include avoiding disagreements, downplaying controversies. Solution-oriented strategies encompass collaboration and compromise, include compromising and searching for innovation. Control strategies include arguing consistently for one's position and using nonverbal messages to emphasize demands.

Of the thirty items in the instrument, twelve of the items measure the usage of nonconfrontation strategies. Statements typical of the twelve items that measure the usage of nonconfrontation strategies include "I shy away from topics that are sources of disputes" and "I steer clear of disagreeable situations."

Eleven of the items measure the usage of solution-oriented strategies. Among them, six of the items measure collaboration. A typical statement includes "I suggest solutions which combine a variety of viewpoints." The other five items measure compromise and the statement "I give in a little on my ideas when my supervisor also gives in" is representative of the statements that measure it.

The remaining seven items measure the use of control strategies to manage and resolve conflict. " I make my opinions known in a disagreement with my supervisor" reflects the nature of the statements that measure the usage of control strategies.

Procedure
The survey questionnaire consisted of a cover sheet explaining the purpose of our study and the 30 statements. The survey questionnaires were then distributed to a convenient sample of 14 people working in an organization. Altogether, we distributed the survey questionnaires to 7 males and 7 females. The respondents were instructed to fill in the questionnaire with respect to the disagreements that they had encountered in a particular task situation with their immediate supervisors. The respondents were then asked to think of a particular situation and indicate how frequently they engage in each of the described behavior on a 7-item Likert type scale. On this 7-item scale, 1 represents always, 2 represents very often, 3 represents often, 4 represents sometimes, 5 represents seldom, 6 represents very seldom, and 7 represents never. Respondents were also asked to indicate their gender at the end of the questionnaire.

All the respondents completed the survey questionnaires, giving a response rate of 100 percent. The surveys were administered and collected within a day.

Data Treatment
The data was collected and keyed into the SPSS software. The total score for each strategy was computed by adding up the scores for all the items measuring each strategy. Because 1 represents always and 7 represents never, low scores for each strategy mean that the strategies are used frequently. Because the number of items measuring each strategy differs, the range of scores for each strategy is also different. The possible range of scores for nonconfrontation strategies is 12 to 84. The possible scores for solution-oriented strategies and control strategies both range from 11 to 77.

Next, the means for each of the three conflict resolution strategies were computed by dividing the total score of each strategy by the number of items measuring each strategy. This was done to enable the comparison of the three conflict strategies, as the number of items measuring each conflict strategy is different. Going by the same principle, the lower the mean score of the strategy, the more frequently the strategy is used to handle conflicts.

Since the main purpose of our study is to find out if gender differences affect the preference for conflict strategy when it comes to managing and resolving conflicts in an organizational setting, we applied the Independent Samples t-Test to the total scores of the three conflict strategies. For our discussion purposes, the level of significance is set at p = .10. We also calculated the mean scores of each conflict strategy for both sexes in order to facilitate the interpretation of our findings.

Results
Independent Samples t-Test
Based on the Levene test for equality of variances under the Independent Samples t-Test, the observed significance level for each of the conflict strategy is greater than .05 level of significance. Therefore, we should not reject the assumption that the variances are equal. Thus, when analyzing our results of the t-test, we referred to the values in the row labeled Equal variances assumed (see Table ).

Nonconfrontation Strategies
The total scores of the respondents for nonconfrontation strategies range from 34 to 61 (see Table ). The mean total scores of the male and female respondents are 50.43 and 49.00 respectively (see Table). Judging from the mean total scores, female respondents used nonconfrontation strategies to resolve conflict more frequently than the male respondents. There is a mean difference of 1.43 at the observed two-tailed significance level of .767 (see Table ). As this observed level of significance is much larger than .10 level of significance, there is no statistically significant difference between the male and female mean scores for nonconfrontation strategies. Therefore, it is likely that gender does not affect the use of nonconfrontation strategies for resolving conflicts.

Solution-oriented Strategies
The total scores of the respondents for solution-oriented strategies range from 30 to 50 (see Table) and the mean total scores of the male and female respondents are calculated to be 36.80 and 39.83 respectively (see Table ). The mean total scores indicate that female respondents were less likely to use solution-oriented strategies to resolve conflicts than the male respondents. There is a mean difference of -4.17 between the male and female scores and the observed two-tailed level of significance of this mean difference is .279 (see Table ). This observed level of significance is too large for us to conclude that there is a statistically significant difference between the male and female scores for solution-oriented strategies. Therefore, it is likely that gender differences do not affect the use of solution-oriented strategies to resolve conflicts despite the mean differences.

Control Strategies
The total scores of the respondents for control strategies range from 19 to 44 (see Table ) and the mean total scores of male and female respondents are 29.86 and 36.14 respectively (see Table ). The scores reflect that the female respondents used control strategies to resolve conflict less frequently than the male respondents. There is a mean difference of -6.29 between the scores for the male and female respondents, observed at .132 level of significance (see Table). As this observed two-tailed level of significance is larger than .10 level of significance, we can conclude that there is no statistically significant difference between the male and female scores for control strategies. Therefore, it is likely that gender has no significant effect on the preference for control strategies.

According to the mean scores of the three conflict strategies, it appears that both the male and female respondents favor the use of solution-oriented strategies, but used control strategies least frequently (see Table ). The comparatively low mean score of 3.4318 for solution-oriented strategies suggests that the respondents were more likely to use solution-oriented strategies or used solution-oriented strategies more frequently to resolve conflicts. The respondents scored a mean of 4.1528 for nonconfrontation strategies. The comparatively high mean score of 4.7143 seems to indicate that the respondents were less likely or used control strategies less frequently to resolve conflicts.

It may be interesting to note that the female respondents had a high mean score of 5.1633 for control strategies compared to the mean score of 4.2653 for the male respondents. This means that the female respondents used control strategies to resolve conflicts less frequently than the male respondents. This mean difference of .898 is the largest mean difference between the scores for male and female respondents (see Table ). However, the results of the earlier t - test indicated that this mean difference is not statistically significant (see Table ).

Discussion
The main purpose of our study is to investigate the effect of gender on the usage of conflict resolution strategies. As mentioned earlier in the literature review, we did a search on the relevant literature and found that there are opposing views on the effects of gender differences on the use of conflict style or strategy. According to Dowd (1991), women tend to adopt cooperative strategies while men tend to be more coercive and forcing when resolving conflicts. Rahim (1983) pointed out that other researchers such as Imler (1980) and Kilmann & Thomas (1977) also found evidence to show that men are more dominating and less compromising than females in conflict situations. A more relevant study done on 303 Singaporeans by Mc Kenna and Richardson in 1995 also supported the findings that gender differences affect conflict-handling behavior. The study revealed that Singaporean males tend to adopt compromising strategies while the females tend to use avoiding strategies.

However, although much research has indicated that gender affects the use of conflict style or strategy, other research has also suggested that the relationship is not so clear. Duane (1989) found that men and women did not differ on their use of collaborative or compromising modes of conflict management.

Our study has shown that gender has no significant effect on how frequent employees use nonconfrontation strategies, solution-oriented strategies, and control strategies to resolve conflicts that occur in an organizational setting. The possible explanations for our findings could be that there is no actual difference in the frequency of usage of the strategies between male and female employees or that the difference is too small to be statistically significant. Based on the findings of our study, it would appear that our results support the findings of past research that the effect of gender differences on conflict-handing behavior is not significant. However, we cannot assume that our findings are valid or representative of the population of interest because of the limitations within our study and the instrument itself.

Limitations and Modifications
The method of our study and the procedure in which we went about conducting our study pose several problems that were made worse by the time and cost constraints.

First, our sample size of 14 was too small to be representative of any population. This is further compounded by the fact that we are interested in studying the effects of gender differences. Thus, we had to stratify the sample into male and female respondents. This gave us only a sample of 7 males and 7 females for our study. In order for a sample to be representative of the population, the sample size must be at least 30. Often, researchers would set the confidence level of the findings and the margin of error allowed prior to determining the sample size. Thus, the researchers should have already taken the confidence level and margin of error into consideration when determining the sample size.

The generalizabilty of our findings also suffers from the fact that our sampling method was far from ideal. Our survey questionnaires were distributed to a convenience sample. This non-probability sampling method is biased and does not ensure that every working individual or at least everyone in the population of interest has an equal chance of being selected to participate in the study. Thus, the findings would not be representative of the population that we are interested in studying in.

The ideal sampling method in our case would be to obtain a comprehensive list of the names of all potential participants in the population that we are interested to study. Next, we should stratify this population into two subsets - males and females. After determining the actual proportion of males to females in our population, we should randomly select the number of male and female participants from the list based on this actual proportion and the sample size required to be representative of the population.

We also need to modify the actual administration of the survey questionnaire when administering them to a larger sample. With a sample size of 14, it was much easier to administer and collect the survey questionnaires. Despite the fact that ours was a self-administered survey questionnaire, we had a high response rate because of the small sample size. However, with a sample size of at least 60 (30 males and 30 females) in order for the findings to be representative, we would have more difficulty getting a high response rate. Thus, we should prepare for more survey questionnaires and administer them to a sample larger than our ideal sample size to safeguard against a low response rate. In this way, the findings for the sample will remain representative of the population that we are studying. We should also prepare for follow-ups on the survey questionnaires in order to elicit a higher response rate.

Modifications also need to be made to the instrument used in our study. For our study, we adopted an instrument - the Organizational Communication Conflict Instrument (OCCI) that has already been tested, refined, and validated. Therefore, any modifications to the instrument should be minimal so as not to affect the validity of the instrument. However, feedback from some of our respondents indicated that the phrasing of certain statements is awkward, making the statements difficult to understand. For example, the statement "I integrate arguments into a new solution from issues raised in a dispute with my supervisor" requires rereading before one can actually understand what the statement means.

We must always bear in mind that the wordings of survey questionnaires should be clear and unambiguous. This is especially important when the potential participants could be of varying educational and literacy levels. Therefore, all statements that appear to be difficult to understand should be modified to improve the respondents' understanding of the sentences. However, it is important that the sentences are not altered in such a way as to affect the meaning of the sentences as this will affect the validity of the instrument. A pilot test can be conducted on a group of participants to test out the questions and the participants selected should be as similar to the participants of the study as possible.

Before administering the survey questionnaires to a larger sample, some other modifications need to be made to the study. In line with the main purpose of our study to test for gender effects, it is preferable that we control as much of the other factors as possible in order to eliminate the effects of other confounding factors. Factors such as age and educational level of the respondents, the personality of each individual i.e. individual differences, the role status and rank of the respondents and occupational differences may all affect conflict-handling behavior.

Unfortunately, this list is not exhaustive and there are many other factors that may account for differences in the usage of conflict resolution strategies. Also, many of these factors may not be easily controlled or accounted for.

As mentioned earlier in the literature review, past research has shown that individual differences in terms of psychological type, personality, or values, affect the use of conflict styles. It was also mentioned that gender is an accurate indicator of the psychological type of a person. Previous research has consistently shown that psychological type correlates with conflict-handling style. Thus, there is a need for us to replicate previous studies in order to increase the validity of our study.

We should be able to address this issue by asking the respondents to fill in another survey questionnaire measuring their psychological type on the Jungian personality dimension.. The findings of this survey questionnaire can then be combined with the finding of the OCCI to establish whether individual and gender differences affect the use of conflict strategies. We can also then establish if there is a relationship among gender, psychological type, and conflict-handling style or strategy.

Another major problem is that the instrument that we chose to use for our survey questionnaire -- the OCCI, has its own limitations. According to King and Miles (1990), the instrument uses the individual and not the relationship or the dynamic of the conflict episode as the unit of measurement. Also, the instrument measures an employee's use of conflict strategies in relation to a specific scenario, in this case, the disagreements that the employee encountered in a particular task situation with the immediate supervisor. This means that the instrument does not take into account the possibility that the context of the conflict or the circumstances in which the conflict takes place may also affect the choice of conflict strategies.

However, the culture or climate of the organization may affect conflict-resolution behavior. The gender stereotypes in an organization may affect the male and female employees' perception of how they should resolve conflicts that occur in the organization. Similarly, the structure of the organization may also create preconceived notions of how employees should resolve conflicts. Thus employees may alter their conflict-resolution behaviors according to these perceptions. Also, the structure of the organization and the nature of its business dealings may affect the types of conflict that would occur in the organization.

Taking into account this inherent limitation, we can perhaps increase the value of our study by narrowing the population of study before we administer the surveys on a larger sample. Ideally, the participants of our study should come from organizations that have similar cultures and climates as their working environments may affect their attitudes towards and their usage of conflict resolution strategies. This would allow us to control the possible effects of these external factors.

Therefore, before we administer the survey, we ought to be very clear about the loopholes and the limitations of the study. We should also conduct sufficient research to determine which are the factors that affect the usage of conflict strategies more than the other factors. Since the main focus of our survey is to study the effect of gender and individual differences across similar organizational setting, we can narrow our focus by selecting a few organizations of similar working environment and having comparable business dealings, for example, local financial institutions. Participants would then be selected from the name lists of these organizations. They would be requested to complete both the survey questionnaires that measure the usage of conflict strategies and their personality with a suitable lapse of time in between the two survey questionnaires. This is to ensure that the respondents would not be affected by the earlier questionnaire that they would have to fill in.

However, the fact that our study relies on the self-report of the respondents with respect to the disagreements that they had encountered in a particular task situation with their immediate supervisor poses another limitation for our study. The accuracy of our survey findings may be affected by the lack of accurate memory recall and self-report from the respondents. In order to tackle this problem, we should include observations of actual conflict resolutions in an organizational setting to obtain a triangulated view for our study. However, this is not taking into consideration the time and cost constraints that will limit the scope of the study. Also, it may not be easy to engage in ethnographic observation in an organization.

Future Directions
According to Gayle (1991), the effect of gender differences on the choice of conflict style or strategy may be rooted in traditional power and status differences between men and women. Rahim (1983) also mentioned that researchers have found that in superior-subordinate communication, individuals are more likely to use the avoiding style with superiors than peers, and more with peers than with subordinates. Conversely, individuals are more likely to use a forcing approaching with subordinates than with peers, and much less likely with superiors. However, the OCCI instructed the respondents to assume a subordinate position. Thus, much valuable information may be lost. Perhaps future research can address this problem by administering three sets of questionnaires with the respondents. The three separate questionnaires can then focus on each of the conflict situation - disagreements with an immediate supervisor, peers and subordinates respectively.

Future research may also focus on the context of the conflicts to determine if conflict- resolution behavior is dependent on the situations and circumstances in which they arose. Another interesting aspect that researchers may want to look into is whether age is an important variable in affecting the use of conflict style or strategy in an organization.

In order to work around the limitation of most instruments that measure conflict management in organizations, researchers may want to conduct similar studies across different kinds of organization and explore the possible effects that different working environments have on the type of conflict strategies employees adopt in different organizations. Comparisons can then be made with studies conducted on similar organizations to obtain a comprehensive picture.

Future research may also want to study whether the relationships of employees within a department and across different departments affect the choice of conflict style or strategy they use to resolve conflicts. The interdependence or expectation of dependence of the employees within or across departments may actually affect their conflict management and thus worthy of future study.

REFERENCES