Back to Journal Cover Page


CURRENT RESEARCH IN SOCIAL PSYCHOLOGY


Volume 9, Number 1
Submitted: July 7, 2003
First Revision: July 9, 2003
Second Revision: July 11, 2003
Third Revision: August 6, 2003
Fourth Revision: August 27, 2003

Accepted: September 2, 2003
Publication Date: September 3, 2003

Macro- and Micro-Social Conditions Affecting Individual Sense of Mattering during a Period of Downsizing

David E. Rohall
Western Illinois University

ABSTRACT

Mattering is a dimension of our self-concept referring to how important we think we are to others. Social interaction is essential to the development of mattering because it provides the opportunity to affirm one's connections with others. I use longitudinal data from a survey of Russian Army officers undergoing organizational downsizing to assess the relative effects of macro-social conditions, measured in terms of distance to urban centers, and micro-social conditions, measured by quantity of social interaction and employment status, on individual sense of mattering. I find that both macro- and micro-level social conditions can impact sense of mattering, albeit in different ways. On the individual level, being unemployed is associated with the greatest loss of mattering over time, probably because the unemployed no longer represent a major source of income support for their families, whereas respondents with more social interaction report higher levels of mattering. On the macro level, respondents living in more remote areas report lower levels of mattering than those closer to cities and towns. Additional findings show that the positive effects of social interaction on mattering are particularly strong in more remote areas.

[1]
---------------
[2]

INTRODUCTION

A recurring issue in military life is the need to relocate from friends and family members (Segal and Harris 1993). In some sense, the military structures the social lives of their service members, taking them outside of their normal social circles and providing a different selection of people with whom to interact. If social conditions help mold the self-concept, it suggests that the military lifestyle may restructure soldiers' self-concept to reflect current social and structural conditions.

Such a lifestyle may have a particularly profound effect on a relatively understudied element of the self-concept called "mattering." Rosenberg and McCullough (1981) introduced the concept of mattering over twenty years ago, referring to the feeling that one makes a difference in other people's lives. This is an important element of social integration but substantively different than traditional integration measures such as social support or anomie. For instance, social support generally refers to the degree to which we feel that we can rely on others to help us, suggesting that other people are important to us. But mattering is a sense that we are important to others, that we "matter" to them. Hence, mattering completes the circle in which we are connected to others and others are connected to us.

My research is designed to explore the social nature of mattering and conditions affecting individual sense of mattering using data from a survey Russian Army officers undergoing organizational downsizing. The advantage of using this sample is that it allows me to examine variations in macro-level social conditions, because respondents were stationed throughout the Russian countryside, as well as micro-level conditions surrounding them. Thus, there is an opportunity to examine direct and interactive effects of both macro- and micro-level conditions on the self-concept.

The Concept of Mattering

Rosenberg and McCullough (1981) define mattering as, "the feeling that others depend on us, are interested in us, are concerned with our fate, or experience us as an ego-extension" (165). In this seminal work, mattering is measured using three items specifically developed for the study, asking how interested adolescents thought their mothers are in them, how important they are to their families, and how interested their parents are in what they had to say. The authors find mattering positively related to self-esteem and negatively related to depression and anxiety, independent of one's self-esteem. Mattering is also negatively related to delinquent behavior: adolescents who feel that they matter more are less likely to engage in delinquent behavior.

[2]
---------------
[3]

The social conditions that influence sense of mattering may stem from social positions and interactions that provide feedback about an individual's relative contribution to a group. For instance, people in high positions, denoting more responsibilities, may have a higher sense of mattering simply because they are more important to more people. Similarly, family members, particularly parents, may have a high sense of mattering because they are legally responsible for the lives of their children. Schieman and Taylor (2001) find that older people tend to report higher levels of mattering, as do people who feel they have more control over their work lives. Their finding makes sense because many people accrue relationships as they age and more autonomy at work implies some sort of leadership role in an organization. In both cases people matter more because they are in a position that matters.

Some research suggests that mattering can result from both the quality and quantity of interaction with other people. For instance, Schieman, Taylor, and Rohall (2002) show that deficits in early childhood paternal warmth to be associated with reduced sense of mattering in adulthood. But simply having access to social interaction can impact sense of mattering too. People who are currently in a close relationship report higher levels of mattering than those who are not in a close relationship (Schieman and Taylor 2001). In this sense, social relationships represent an opportunity to matter; simply being around people implies a social connection. From a social exchange perspective, there must be something of value to give in a relationship in order for it to exist. Thus, simply being with other people may give people a sense of value. Given this research and theory, I predict that mattering will be positively related to levels of social interaction: individuals who interact more often will report higher levels of mattering.

Social Conditions Affecting the Self-Concept

Separation is a common aspect of military life. Soldiers are often asked to leave their families for days and months at a time (see Rohall, Segal, and Segal 1999). This type of separation affects personal well-being in many ways, including higher levels of anxiety and depression (Black 1993; Kelly 1994). Another major form of separation that affects military members is separation from home and lifestyle. That is, soldiers and their families are often asked to leave long-standing relationships to be stationed somewhere else (Segal and Harris 1993).

[3]
---------------
[4]

Separations in the Russian context can have a more profound impact on soldiers simply by virtue of the Russian landscape. While the Russian population is at least third less than the U.S. population, it has about twice the landmass to protect. Much of that land, especially in central (e.g., Siberia) and far eastern portions, is desolate with little access to transportation or infrastructure. Thus, bases in these areas provide little relief for Russian soldiers and their families.

While no research on macro-social conditions affecting mattering exist, a number of studies examine how various social conditions may affect other aspects of the self-concept, notably self-esteem, a correlate of mattering. For instance, on the meso-level, Wiltfang and Scarbecz (1990) find that neighborhood unemployment has a strong, negative effect on self-esteem. Further, Abu-Saad (1999) finds that adolescents with better relationships with parents and peers and those living in more urban environments report higher levels of self-esteem than those living in more remote villages.

Social conditions can impact the individual by providing some indication of social position and relative worth in a given environment. But these reflections are largely mediated through direct social interaction. In one study, Whitbeck and his colleagues (1991) find that economic hardship affects children's self-esteem by decreasing social support and involvement. Similarly, Gecas and Seff (1990) find that the effects of social structural variables on self-esteem depend on the extent in which these variables operate: when social structural variables are more salient, their impact on the self-concept is stronger. In this sense, the self-concept is somewhat pliable, providing different information from which to assess one's self-esteem under different social conditions (Bohrnstedt and Fisher, 1986).

These findings suggest that the relationship of macro- and micro-level conditions may impact mattering in two ways: macro-level conditions can have direct effects on mattering as well as interactive effects with micro-level conditions, making micro-level conditions more or less salient to people. Social structural conditions are particularly relevant for soldiers who are often stationed far away from traditional sources of social support like friends and family. Given the role of social conditions on the self-concept more generally, it is predicted that individuals living in more remote places will report lower levels of mattering than those less isolated places. However, it also predicted that the effects of individual-level social interaction on mattering will vary by location. Specifically, the effects of social interaction on mattering will be stronger in more remote areas than in less remote areas because that interaction will be more important in maintaining mattering.

[4]
---------------
[5]

Downsizing and the Self-Concept

Most research on the effects of organizational downsizing on the self-concept focus on its impact on personal well-being, such as depression and anxiety. And most downsizing research is done in civilian contexts. Though there is some evidence that organizational downsizing produces distress among survivors, the preponderance of evidence is that those who are asked to leave an employer do worse than the survivors. In terms of mental health, those leaving a job as a result of downsizing show greater signs of anxiety, depression, hostility, somatization, and alcohol abuse than survivors of organizational downsizing (Hamilton, Hoffman, Broman, and Rauma 1993; Hoffman, Carpentier-Alting, Thomas, Hamilton, and Broman 1991; Kessler, Turner, and House 1988; Liem 1981; Moen 1980).

While there is some research that suggests job loss may also reduce one's sense of self-esteem, the findings are mixed. Dooley and Prause (1995) find that unemployment is negatively related to self-esteem. Further, longer durations of unemployment are associated with lower self-esteem (Sheeran and McCarthy 1990). In another study, Sheeran and Abraham (1994) argue that most of the effects of unemployment on the self-concept are largely mediated via reflected appraisals from friends, family, and other people. Under this schema, life events affect how people see and interact with us. Kessler and McCleod (1984) argue that income loss is particularly negative for men compared to women because of the importance society places on work for men. Job loss reduces one's sense of value in society and in the family directly, via loss of income. While no one has studied the impact of unemployment on mattering, Schieman and Taylor (2001) do find a positive relationship between income and mattering. Loss of a job not only represents loss of income but loss of position. Thus, it is predicted that job loss will have a direct, negative effect on mattering because employment represents a source mattering to family, via income support, and loss of position in the organization.

The larger social context can also have an impact on how people react to job loss and downsizing. Most research in this area focuses on the effects of larger economic conditions on personal sense of well-being. For instance, Dooley, Catalano, and Rook (1988) find direct effects of community-level unemployment rates on individual distress but no interaction between job loss and larger social conditions on distress. More recent research shows that community-level unemployment rates are not directly related to depression levels but affect distress indirectly by increasing the risk of becoming unemployed (Dooley, Catalano, and Wilson 1994). However, the focus on this paper is not on the impacts of economic conditions on mattering per se but the social conditions surrounding the relationship between downsizing and sense of mattering. But given the previous research suggesting the possibility of both direct and indirect effects of macro-level conditions on individual well-being during a period of downsizing, it is predicted that the negative effects of job loss on mattering will be heightened in more remote areas, where there are fewer social resources with which to cope with unemployment.

[5]
---------------
[6]

METHOD

Sample

The Russian Army has been in the process of downsizing since the late 1980s. Researchers from the University of Maryland, the University of Michigan, and other locations supervised a two-stage panel survey of Russian Army undergoing downsizing in the mid to late 1990s. They only sampled officers because Russian enlisted personnel are largely drafted from the general population, making the separation experience quite different for this group. Army lists of 1800 officers who would be leaving within 3-6 months were randomly sampled, clustered geographically by region, nested in 58 bases. The seven military regions or districts in Russia range from very rural areas like the Far East to highly urban areas such as Moscow. Twenty-four percent of the respondents came from the St. Petersburg (Leningrad) district, followed by the Volga district representing 20% of the sample, Kaliningrad with 16% of the sample, Trans-Baikal 14%, Far East 13%, Moscow 9%, and Ural 3% (differences due to rounding).

The interview schedule included a number of questions about the downsizing experience, including job-searching strategies and the cognitive impacts of downsizing, including measures of the self-concept, life satisfaction, distress levels, organizational commitment, drinking and substance abuse, financial hardships, among others. The instruments were written in English, translated to Russian in Moscow, and back-translated into English in Moscow and in the U.S., where the two versions were compared to resolve any discrepancies. On average, interviews lasted about 60-67 minutes.

Army psychological services personnel administered the first set of surveys in late 1995 with a group of Russian officers about half of whom were slated to leave service as a result of organizational downsizing, yielding a response rate of 90% or 1,798 interviews. They interviewed the officers again in the spring and summer of 1997, over a year after the first set of interviews, producing a response of 85% (1,536 respondents).

Since this study compares those who have actually left the military with those who are still in service, I will focus on the responses from the second wave of data collection, though some first wave data are included as control variables, to predict changes in variables. This study includes officers who remained in service after the downsizing event (wave 2) and those who left and either found employment or are currently unemployed and actively looking for employment because the latter group represents those who were involuntarily separated from service. I exclude those officers who left service and are on paid or unpaid leave (n=8) or did not respond to the question regarding to their job status (n=19) and those who are unemployed and not looking for work (n=14) or who did not respond to the question about whether they are looking for work (n=19), reducing the sample size to 1,476 in 49 different bases.

[6]
---------------
[7]

Dependent Variable

My measure of mattering is derived from the global mattering scale developed by Marcus and Rosenberg (1987) that includes the following items:

How important do you feel you are to other people?

How much do you feel other people pay attention to you?

How much would other people miss you if you went away?

How interested are people generally in what you have to say?

How much do other people depend on you?

Responses for all items ranged from 1 "A Lot" to 4 "Not At All." These items are coded so that higher scores are associated with higher levels of mattering, producing a Cronbach's alpha reliability coefficient of .85 for officers' mattering with a mean of 2.95 (see Table 1).

Table 1: Sample Characteristics by Region*

No. of Bases

N

Mattering

Distance to City (km)

Social Interaction

Employment Status (1=unemployed)

Military Region

Mean (sd)

Mean (sd)

Mean (sd)

Mean (sd)

Kaliningrad

6

240

2.76 (.77)

18 (25)

3.66 (1.22)

.10 (.31)

St. Petersburg

9

354

2.87 (.49)

45 (68)

3.27 (1.12)

.05 (.23)

Moscow

3

140

3.06 (.51)

10 (18)

3.54 (1.30)

.09 (.29)

Volga

6

293

3.17 (.58)

19 (15)

3.46 (1.00)

.12 (.33)

Urals

5

41

2.93 (.90)

9 (7)

3.68 ( .92)

.12 (.33)

Trans Baikal

10

212

2.86 (.58)

123 (105)

3.82 (1.41)

.09 (.28)

Far East

10

196

3.01 (.53)

17 (31)

3.59 (1.33)

.05 (.21)

Total

49

1476

2.95 (.61)

39 (66)

3.53 (1.21)

.08 (.28)

* All models tested using one-way ANOVA analysis; Mattering (F=14.185, p<.001), Distance (F=110.076, p<.001), Interaction (F=5.435, p<.001), Unemployment (F=2.495, p<.05)

[7]
---------------
[8]

Independent Variables

Micro-level social conditions predicted to impact sense of mattering include levels of social interaction with friends, family, and neighbors and employment status. Levels of social interaction are measured by responding to the question: "How often do you get together with neighbors, friends or relatives, for example, go out together or visit each other?" Responses range from 1 "several times a week" to 6 "never." I reverse coded the item so that higher numbers are associated with more interaction, producing a mean of 3.53 (Table 1).

Employment status may have a large effect on sense of mattering simply because income from work represents a major source of supporting other people, a sense of contributing to the physical well-being of others. Of the 1,476 respondents in this study, 1,352 or 92% either remained in service after downsizing or found new work. Thus, only 124 (8%) are unemployed leavers.

The primary macro-level condition predicted to affect personal sense of mattering is distance to the nearest city. Remoteness to cities means more than reducing access to physical survival but also to communication resources and sources of social interaction. Distance to nearest city refers to the distance the respondent is to any city and is measured in kilometers, based on the 49 sample points. Base-level measurements are provided by the International Center for Human Values (1997) in Moscow, Russia. The macro-level data are based on government statistics that were provided between the first and second wave of data collection. The average distance to the nearest city ranges from 0 kilometers (i.e., within a city) to 400 kilometers with a mean distance of 39 kilometers (using base-level data).

Age and marital status (married or not) are included in the analyses as control variables, controlling for individual-level conditions that may affect sense of mattering (see Schieman and Taylor 2001). It is important to note that age can also be used as a proxy for rank and time in service; older soldiers tend to have both higher rank and more time in service. Age is used instead of rank because rank is no longer relevant for officers who left service. The average age of this sample is 34 and almost all are married (96%). Social interaction at wave 1 is also controlled in order to test if changes in social interaction are associated with current sense of mattering.

[8]
---------------
[9]

Analysis

In addition to a review of the basic relationships between model variables, hierarchical linear modeling (HLM) will be conducted to determine if the larger social conditions impacts individual-level sense of mattering and its interaction with micro-level conditions. The HLM technique is designed to distinguish group-level measures from individual-level measures when partitioning the variance in a regression model. Traditional forms of regression such as (OLS) regression tend to underestimate the standard errors associated with group-level variables because OLS regression assumes statistical independence. This assumption is violated when group-level measures are attached to individual-level records because those scores are the same for each individual in that group; hence, they are not independent (Hirsch and Schumacher 1992).

HLM calculates a separate regression slope for each individual-level variable within each group-level variable (Bryk and Raudenbush 1992). Thus, HLM provides the relative significance of individual and group-level variables on the outcome variable. Two components of an HLM include a model on the individual level and another on the group level. The N on the group level is the number of group-level units in the analysis. This analysis uses base-level measures (n=49) as the group measure rather than the regional measure because there are only seven regions, not enough for advanced analysis. The result of this technique is to examine direct effects of individual- and macro-level variables on mattering, controlling for each other, as well as an assessment of the interaction between these levels of measurement.

RESULTS

Macro-Level Variations in Mattering

The first set of analyses is designed to determine if there are any variations in mattering across locations. This study includes officers and former officers in 49 locations across the Russian countryside, under varying conditions. Some live within cities while others live in very remote locations, sometimes as much as 400 kilometers from a major town or city.

[9]
---------------
[10]

Table 1 provides data on average mattering scores, distance to nearest city, social interaction levels, and unemployment rates on the regional level. These data show that areas with the highest average mattering levels are located in the western and middle parts of Russia, in the Volga district (3.17) and Moscow (3.06), Russia's largest city. However, social interaction is highest in one of the districts in the eastern portion Russia, where there are fewer urban centers — where they probably need it the most. For instance, respondents in the Trans-Baikal region, on average, report the highest levels of interaction (3.82) and the furthest distance to a city (123 kilometers). However, these findings are averages of base-level scores within each respective region; limitations on information about specific locations were a result of security concerns. Hence, it is important to be aware that the summary data in Table 1 obscures the conditions of individual bases within each region. However, all advanced, HLM analysis will utilize base-level data when discussing macro-level effects, though I cannot determine their exact location of a particular base within a specific region.

City Distance, Social Interaction, Employment Status, and Sense of Mattering

Mattering is correlated with all model variables in the predicted direction: as distance to the nearest city goes up, sense of mattering goes down (r=-.11, p<.001), supporting the second hypothesis (Table 2). Conversely, sense of mattering is positively associated with social interaction (r=.07, p<.01): the more time people spend with other people, the higher their sense of mattering, supporting the first hypothesis. Unemployed people show lower levels of mattering than those who are employed (r=-.13, p<001), supporting the third hypothesis.

Table 2: Correlation of Mattering and Other Model Variables

 

Mattering

City

Distance

Social

Interaction

Unemploy-ment

Marital

Status

Age

Mattering

----

         

City Distance

-.11***
(1453)

----

       

Social Interaction

.07**
(1384)

-.04
(1405)

----

     

Unemployment (1=unemployed)

-.13***
(1453)

.05
(1476)

.02
(1405)

----

   

Marital Status

.10***
(1451)

-.01
(1473)

-.13***
(1402)

-.03
(1473)

----

 

Age

.07**
(1446)

-.05
(1469)

-.14***
(1398)

-.10***
(1469)

.15***
(1466)

----

* p. < .05, ** p. < .01, *** p. < .001

[10]
---------------
[11]

Direct and Interactive Effects of Macro- and Micro-Level Conditions

While traditional statistics suggest that mattering is related to macro- and micro-level conditions, they do not provide evidence of interactive effects. Further, the effects of macro-level conditions may be misleading because base-level data are attached to individual-level records. HLM analysis separates these levels of analyses when determining the direct and indirect relationships between them.

I predicted that macro-level access to social resources, measured in terms of distance to the nearest city, would have a small, direct effect on mattering but that some of its effects would be felt through individual-level processes, notably by structuring social interactions. Initial findings continue to support the hypothesis that there should be variations in levels of mattering based on remoteness of location (chi-square=102.6, p<.001): average levels of mattering vary by base location (Table 3). The next stage is to determine the conditions that affect mattering and spatial variations in mattering.

Table 3: HLM Model of Mattering, Social Interaction, City Distance, and Employment Status

Coefficient

se

t-ratio

p-value

Random Effects

Average base mean score (Intercept)

2.9459

.0351

83.949

.000

Distance to Nearest City

-.0095

.0005

-2.015

.049

Fixed Effects

Level of Social Interaction (Wave 2)

.0477

.0137

3.489

.001

Distance to Nearest City

.0004

.0002

2.070

.044

Employment Status (1=Unemployed)

-.3349

.0910

-3.677

.001

Distance to Nearest City

.0002

.0012

0.195

.978

Marital Status (W2, 1=Married)

.2334

.0854

2.732

.007

Age (W1)

.0084

.0025

3.390

.001

Level of Social Interaction (Wave 1)

.0013

.0122

0.105

.917

Chi-Square

Standard Deviation

Variance Component

df

Chi-Square

p-value

Base Mean (Level-2 Effect)

.1935

.0375

28

102.567

.000

Level of Social Interaction (Wave 2)

.3648

.1331

28

55.189

.002

Employment Status (1=Unemployed)

.0256

.0006

28

23.448

>.500

Level-1 Effect

.5613

.3150

[11]
---------------
[12]

The second hypothesis, that mattering would be negatively associated with the distance to the nearest city, also has some support, even after controlling for background characteristics usually associated with mattering. The random effects model of the HLM statistics focuses on the effects of base-level variations in city distance on personal sense of mattering (b=-.001, p<.05). Hence, for every kilometer in distance, there is a .001 unit loss of mattering. Put differently, for every 600 miles in the distance to the nearest city, there is a one unit decrease in an individual's sense of mattering to other people.

Two additional statistics focus on the direct effects of social interaction on mattering as well as the direct effects of job loss on mattering (Table 3, Fixed Effects Model). Supporting the first hypothesis, levels of social interaction are positively related to personal sense of mattering (b=.05, p<.001). Further, the fourth hypothesis, that being unemployed would be associated with lower sense of mattering is also supported (b=-.33, p<.001).

The final set of analyses is designed to determine the interaction between micro- and macro-level conditions, predicting that social interaction and downsizing status should have a stronger impact on mattering in more remote locations. The third hypothesis is generally supported: the effects of social interaction on mattering increases in more remote locations (b=.0004, p<.05), however, this effect is constant across locations. The fifth hypothesis about the interaction of job loss and location on mattering is not supported: the effects of unemployment on mattering do not seem to interact with location at all.

DISCUSSION

Studying the relationship between macro- and micro-level conditions can be challenging because it is difficult to assess the relevant structural conditions surrounding each individual. For instance, Census data places individuals into tracts that may or may not correlate with the lived experience of the individuals in that tract. The military context allows researchers to do this kind of analysis because of the limited span of conditions surrounding each military base. The conditions surrounding those areas, in this case the Russian military, then gives us a glimpse of the macro-level environment in which current and former officers in the Russian Army lived.

[12]
---------------
[13]

The study of mattering in this arena is also useful because mattering is derived from the social conditions around us. Mattering is a concept that Rosenberg developed to study the role of the individual in society. It serves as a natural link between social conditions and the self-concept because it is implicitly social in nature. It measures individuals' sense of how important they feel in their social world. Changing social conditions provide different feedback about one's role in their social world. In this sense, social conditions provide information from which our friends and family make reflected appraisals, the mechanism by which self-esteem is affected (see Sheeran and Abraham 1994). This study shows that just spending time with other people produces higher levels of mattering.

Additional work needs to be done to determine if quality of interaction would interact with the quantity of interaction on mattering. Some research would suggest that there are such qualifications. Shieman and Taylor (2001) find that better work conditions, in terms of jobs that allow more control over work, are associated with greater sense of mattering. However, this finding does not reflect the quality of relationships per se, unless the job reflects the worker's relationship to their employer. These findings may simply reflect workers' level of responsibility in their jobs. One way to examine this issue would be to look at the effects of social interaction and add variables regarding the quality of that interaction. Mediating effects can be established if the quality of interaction reduces the effects of the quantity of interaction.

Understanding the link between macro- and micro-level conditions on individuals' sense of self can be challenging both because individual-level conditions may vary by location as well as the macro-level conditions. My study shows that both of these occur at the same time. There are small, direct effects of macro-level conditions, in this case, remoteness represented by distance to the nearest city. Living further away from people reduces sense of mattering, ostensibly because there are fewer opportunities to matter in people's lives. But these effects are quite small compared to its interactive effects on social interaction and mattering. People living further from cities report lower sense of mattering, particularly if they have little social interaction. Hence, social interaction is more important to sense of mattering when there are fewer social resources. In some sense, the few relationships available make each relationship more important as a source of appraisal.

[13]
---------------
[14]

All of these findings occur amidst the backdrop of organizational downsizing. Downsizing can affect people through a number of different relationships, including one's family and work relationships. Unlike their working counterparts in this sample, who are more distressed (see Rohall, Hamilton, Segal, and Segal, 2001), the unemployed group report lower levels of mattering. Losing a job immediately separates people from their colleagues and their employer; clearly one does not have the opportunity to matter to these groups after leaving the job. In this study, unemployment has a direct, negative effect on individuals' sense of mattering but the effects of unemployment did not interact with social structural conditions.

The challenge for future research in this area is to find additional micro- and macro- level factors affecting sense of mattering. My study finds that social interaction is important to mattering but it does not explain all of the variation in mattering. That is, mattering continues to vary by location after controlling for both macro- and micro-level conditions. More information about the breadth and quality of interaction, as well as the social conditions structuring these interactions, may help to explain this variation.

REFERENCES

Abu-Saad, I. 1999. Self-esteem among arab adolescents in Israel. Journal of Social Psychology, 139, 4, 479-487.

Black, William G., Jr. 1993. Military Induced Family Separation: A Stress Reduction. Social Intervention Work, 38, 3, 273-280.

Bohrnstedt, G. W. and Fisher, G. A. 1986. The effects of recalled childhood and adolescent relationships compared to current role performances on young adults' affective functioning. Social Psychology Quarterly, 49, 1, 19-32.

Bryk, A. S. and Raudenbush, S. W. 1992. Hierarchical Models: Applications and Data Analysis Methods. Sage Publications: Newbury Park, London, and New Delhi.

Dooley, D. and Prause, J. 1995. Effect of unemployment on school leavers' self-esteem. Journal of Occupational and Organizational Psychology, 68, 3, 177-192.

Dooley, D., Catalano, R. and Rook, K. S. 1988. Personal and aggregate unemployment and psychological symptoms. Journal of Social Issues, 44, 4, 107-123.

Dooley, D., Catalano, R. and Wilson, G. 1994. Depression and unemployment: Panel findings from the epidemiologic catchment area study. American Journal of Community Psychology, 22, 6, 745-765.

[14]
---------------
[15]

Gecas, V. and Seff, M. 1990. Social class and self-esteem: Psychological Centrality, Compensation, and the Relative Effects of Work and Home. Social Psychology Quarterly, 53, 2, 165-173.

Hamilton, V. L., Hoffman, W. S., Broman, C. L., and Rauma, D. 1993. Unemployment, distress, and coping: A panel study of autoworkers. Journal of Personality and Social Psychology, 65, 234-247.

Hirsch, B. T. and Schumacher, E. J. 1992. Labor, earnings, and the racial composition of jobs. The Journal of Human Relations, 27, 602-628.

Hoffman, W. S., Carpentier-Alting, P., Thomas, D., Hamilton, V. L., and Broman, C .L. 1991. Initial impact of plant closings on automobile workers and their families. The Journal of Contemporary Human Services, February, 103-107.

International Center For Human Values. 1997. The Process of the Armed Forces Reduction In the Russian Federation. Unpublished Manuscript. Moscow.

Kelly, M.L. 1994. The Effects of Military-Induced Separation on Family Factors and Child Behavior. American Journal of Orthopsychiatry 64, 7, 103-111.

Kessler, R. C. and McCleod, J. D. 1984. Sex differences in vulnerability to undesirable life events. American Sociological Review, 49, 5, 620-631.

Kessler, R. C., Turner, J. B., and House, J. S. 1988. Effects of unemployment on health in a community survey: Main, modifying, and mediating effects. Journal of Social Issues, 44, 69-85.

Liem, R. 1981. Unemployment and mental health implications for human service policy. Policy Studies Journal, 10,2, 350-364.

Marcus, F. M. and Rosenberg, M. 1987. Mattering: Its measurement and Significance in everyday life. Paper presented at the 1987 Eastern Sociological Society Meetings.

[15]
---------------
[16]

Moen, P. 1980. Measuring unemployment: Family considerations. Human Relations, 33, 3, 183-192

Rohall, D. E., Hamilton, V. L., Segal, D. R., & Segal, M. W. 2001. Downsizing the Russian Army: Quality of Life and Mental Health Consequences for Former Organizational Members, Survivors and Spouses. Journal of Political and Military Sociology, 29, 73-91.

Rohall, David E., Mady Wechsler Segal, and David R. Segal. 1999. Examining the Importance of Organizational Supports on Family Adjustment to Army Life in a Period of Increasing Separation. Journal of Political and Military Sociology, 27(1): 49-65.

Rosenberg, M. and McCullough, B. C. 1981. Mattering: Inferred significance and mental health among adolescents. Research in Community and Mental Health, 2, 163-182.

Schieman, S. and Taylor, J. 2001. Statuses, roles, and the sense of mattering. Sociological Perspectives, 44, 4, 469-484.

Schieman, S., Taylor, J. and Rohall, D. 2002. Recollections of Empathetic Parent-Child Bonds, Interpersonal Style, and the Sense of Mattering in Adulthood. Paper presented at the Southern Sociological Society meetings in New Orleans.

Segal, Mady W. and Jesse J. Harris. 1993. What We Know About Army Families. Alexandria: Army Research Institute for the Behavioral and Social Sciences. Special Report 21.

Sheeran, P. and Abraham, C. 1994. Unemployment and Self-Conception: A Symbolic Interactionist Perspective. Journal of Community and Applied Social Psychology, 4, 2, 115-129.

Sheeran, P. and McCarthy E. 1990. The impact of unemployment upon self-conception: Evaluation, affection, consistency and involvement dimensions. Social Behaviour, 5, 5, 351-359.

Whiltfang, G. and Scarbecz, M. 1990. Social class and adolescents' self-esteem: Another look. Social Psychology Quarterly, 53, 2, 174-183.

Whitbeck, L. B., Simons, R. L., Conger, R. D., Lorenz, F. O., Huck, S., Elder, G. H. 1991. Family economic hardship, parental support, and adolescent self-esteem. Social Psychological Quarterly, 54, 4, 353-363.

[16]
---------------
[17]

AUTHOR NOTE & ACKNOWLEDGEMENTS

This research was supported in part by the National Science Foundation (Sociology Program) under grants #SBR-9402212, 9411755, and 9601760 and the U.S. Army Research Institute for the Behavioral and Social Sciences under contracts DASW0195K005 and DASW0100K0016. The views expressed here are those of the author, not necessarily those of the National Science Foundation, the Department of Defense, or the U.S. government. I would like to thank Lee Hamilton and John Reynolds for all of their support in developing this research. Details about the data and survey instruments can be obtained by contacting the author.

AUTHOR BIOGRAPHY

David Rohall is an assistant professor of Sociology at Western Illinois University. He currently studies the impact of social structure on individual thoughts, feeling, and behaviors in both civilian and military contexts. His e-mail address is: de-rohall@wiu.edu.

[17]
---------------
[18]