Immigrants bring many advantages to the host country, for example, according to Makabe 1981, Japanese immigrants play an important role in the development of the coffee and cotton industries in Brazil. Also in the United States, in the metropolitan areas, a wide variety of customers are served by immigrant shopkeepers, restaurants and providers of personal and professional services. Also in the manufacturing industries, especially in the garment industry, business owned by immigrant minority are important. Waldinger (1986) states that much of the New York City garment industry has passed from Jewish, Italian, and Hispanic ownership to Chinese and Korean ownership and most garment shops in greater Los Angeles are owned by Vietnamese and Korean immigrants.
There are other advantages from immigration because it can expand the labor supply and the pool of financial capital, to increase the diversity of organizational forms and the growth of ethnic markets.
Other studies that the family can be viewed as an important source of sefl- employment for immigrants because it is a net work of obligations that embodies the social, economic and cultural investments. In other words, the family can be viewed as a social network which can be effectively harnessed to achieve collective goals (Coleman 1988; Fernandez-Kelly and Garcia 1990; Ferree 1979; Hamilton and Kao 1990; Kibria 1994; Perez 1986).
In other words, it can be referred that the family comprises a social network of immigrants that can help to increase the income at individual level
Because of the fact that there is diveristy in the study of the impact of immigrants for the development of locality, Tubadji and Nijkamp (2014) conducted a research on the relationship between human capital and cultural capital in the context of local labour market productivity.
So far the literature there is an acceptance that human activity in a locality can affect that local development, and not much attention is paid to the effects of immigrants on that local development. Researchers have been focused on the different characteristics of immigrants such as education level or the local diversity that they create to explain the their effect on productivity. However, the impact of immigrants on the economic development is not one way, because it does not only depend on the quality and activity of the immigrants themselves. The reason is that the immigrants and the local human capital encourter and interact on market competition and in the social field is not at all standard mechanics, but highly socially and culturally sensitive process. Therefore, the dynamics of the cultural encounter between locals and immigrants should also been taken into account to explain the immigrant effect on regional economics.
When immigrants enter the local job market competition, there are objective and subjective criterias such as creative vs routine jobs (Bowles and Gintis, 1975). However, according to Nijkamp et al (2011), the objective criteria are not equal because different immigrants groups such as EU and on EU origin face different legal entry for the labour market and hence have different bargaining power on the job market. In addition, cultural capital is also another barrier, according to Bourdieu (1986), cultural background helps or hinders the promotion of an immigrant into a society besides his or her actual level of skills. As a result, immigrants who have the right cultural capital will intergrate into the society more successful than workers with less powerful cultural capital.
Therefore, the cutural capital of the locals is more powerful than that of the immigrants and the creavity from locals will be more likely to be institutionalized as a new set of rules than from immigrants. Moreover, according to Bourdieu (1973) and Passeron (1977), the locals will be promoted over the newcomers to ensure their employment and individual welfare. However, this fact results in the reallocating highly skilled but non creative local work force to creative occupations, while creative immigrants (with both high or low skills) can be allocated by the cultural mechanism to routine occupations. This is due to the cultural bias which serves for the benefits of local population. As a result, the potential positive impact of the creative part of the immigrants on the local productivity can be diminished in the long run in exchange for the welfare outcome for the locals in the short run.
In addition, in the long run, both locals and foreigners in the locality become worse off because of the rational choice of the employers whieh prefer creative locals with cultural background, better social skills and better knowlege of institutions.
The authors, Tubadji – Nijkamp (2015) study the impact of cultural gravity which is the mechanism of the cultural impact on the geographic concentration and utilization of human capital in a locality on both the geographic concentration and productivity of human capital from immigrants in 15 countries in the EU region. Cultural capital is defined as the bond with local culture of a person or an entity which is unique for each different person or group of people and each incoming flow of immigrants for a particular locality. Both the local and the immigrants meet in the locality which creates diverse culture in which its effects depends on the presence or absense of conflict of the divergent attitudes, which offers the concept of cultural distance.
By definition, cultural distance means as reversed cultural similarity or proximity or as the sociopsychological difference between localities. Cultural distance can be able to have an impact on only financial decision making but also on human decision making which includes both occupational and migration choices. The Weberian tradition states that culture can impact local investment decisions by occputational choice. According to Florida (2002), Florida and Mellander (2010), local milieu which is the local value system have effects on the concentration of creative workers in a locality. Boeri and Brücker (2005); Ottaviano and Peri (2005) states that local milieu can represent a facilitator or a barrier ( according to Borjas 1994, 1995, 2003; Collier 2001; Angrist and Kugler 2003) to economic development.
Cultural distance can also be defined as a collection of incompatible or diverse aspects between the culture of the locals and that of the immigrants or foreigners. Cultural distance can also be viewed as the social incentive or the social restriction to migration. Saying in another word, this incompatibility between cultures causes a cultural risk which decreases the attractiveness of a locality to foreigners, such as immigrants or investors.
In their research, Tubadji – Nijkamp (2014), it is proved that local cultural capital can determine local productivity by two ways, which are cultural capital attracts foreign human capital and determines the use of different groups of human capital in the locality in an effective manner. Florida (2002a, 2002b, 2005) states that culture has been investigated as a critical factor that generates flows of creative human capital. Also, Axelrod (1997); Williamson (2000); Hofstede (2001); Tiebout (1956) states that, besides economic incentives, amenities, and goods provided, culture openness plays an determinant role in the local development in which culture openess is a driving force for the movement of people and their interactions and
The finding by Tubadji – Nijkamp (2015) is consistent with others by Florida (2002) and Moller and Tubadji (2009) who states that the more open cultural capital can attract immigrants and redistribute the human capital around this area. In addition, living culture is a work environment criterion, which can affect the migration choice and is exercised by the immigrants who can have higher bargaining power on the market. Besides, local cultural capital also influence the allocation of foreign human capital to tasks (and jobs) locally. Therefore, immigrants have different effect in different localities, especifically, when cultural capital eliminates the natural allocation of creative people to creative jobs and trades off creativity for high skills, then incoming human capital can have negative effects on some localities. In addition, the “cultural gravity” effect applies not only to the concentration of immigrants but also to the concentration of other forms of human capital such ás productive capital.
Therefore, the inability to adjust to the immigrants’ cultural preferences leads to the negative effects for immigrant productivity and of the local productivity.
In short, it is ecpected that there is a negative relationship between immigration aversion and social capitall. And immigrants can lead to higher income through self- employment created by family, a network of obligations that embodies the social, economic and cultural investment which can be regarded as a special form of social capital. In addition, it is figured out that the local productivity of a region can be increased if that cultural capital in that region supports the immigrants and allocates jobs to them based on their actual levels of skills When the culture of a region is open, it is expected that that region can enjoy higher productivity contributed by immigrants, therefore leading to higher incomes.
4. The negative effect of income on social capital
Recently, there have been a lot of research which studied the relationship between income inequality and generalized trust and most of them found that at least on the national level that relationship is negative, which means when income inequality is high, the generalized trust is low. (Leigh 2006; Rothstein and Uslaner 2005; Jordahl 2007; Gustavsson and Jordahl 2006; Uslaner and Brown 2005). That fidinng was also found by Wilkinson and Pickett (2011), Putnam (2007). The reason is that in societies where the inequality is high, there are big differences between individuals so that they rarely feel connected and trust each other. It is argued that when inequality is high, people at the two ends of the income distribution, the top and the bottom, will not regard themselves as having a shared fate, said Uslaner and Brown (2005). Therefore, they are less likely to trust people from different backgrounds. Furthermore, they said that “where inequality is high, people will be less likely to believe that the future looks bright, and they will have even fewer reasons to believe that they are the masters of their own fates”.
Inequality perception and stratification effect can also explained the above relationship. First of all, a person who regard income equality in society as high or unfair are more likely to identify less with people of other incomes. Pickett (2009) states that the less privileged can feel envy and jealousy. Moreover, the people who perceive high inequality might have low egalitarian values, leading to lower social trust (Uslaner, 2005, Gustavsson and Jordahl 2006).
Secondly, society which are defined as being stratified as when the rich and the poor live in different neighbourhoods, rarely meet each other and therefore have low trust in each other. Stratification effects hence imply larger distances between social groups, making contact less feasible. In addition, highly stratified society are more closed, leading to the fact that people from different strata seldomly meet each other. This is also in line with many recent studies which figured out that the higher the income inequalities in Western societies, the higher the social boundaries (Elgar and Aitken 2010; Wilkinson and Pickett 2009). On the other hand, stratification effects are not only limited to social contacts but also to resources. The reason is thar in more unequal societies, resources play an important role in getting a job or having the feeling of success which have been proven to affect social trust (Delhey and Newton, 2005). According to the ‘neo-material’ theory, (Lynch et al. 2000), resources are needed to achieve desirable outcomes such as investment in human, physical, health, and social infrastructure. The negative effect of the lack of resources caused by income inequality leads to such underinvestments on both individual and systematic level. In this view, the effect of inequality on trust occurs because of a disparity in the availability of resources.
Stratification effects can also happen through more spurious mechanisms such as in a society where high incomes can raise political influence which has an endogenous relation with social trust (e.g. Newton 2007). Therefore, the higher the income inequality, the less generalized trust through an increase of inequality in political efficacy. Moreover, stratification effects also means that an increase in income inequality leads to the increase of the importance of income as a social stratifier.
With regards to perceived inequality effects, which in contrast to the stratifi cation hypothesis, it is not the actual distribution of resources, but it is the extent to which the distribution is perceived. The effect imply that an increase in income inequality results in an increase of people’s perception of the degree of income inequality, which subsequently has a depressing effect on trust.
Based on the literature above, the theoretical model is built with the expected signs as follows
II.Data collection and Methodology:
1. Data collection:
The data was taken from the European Values Survey which is the most comprehensive survey on European people’s values, beliefs, preferences, attitudes and opionions on family, work, religion, politics and society. It is a large- scale, cross- national and the most comprehensive research which is repeated every nine years in an increasing number of countries all over Europe, the United States and Canada.
In the late 1970s, The European Value Survey (EVS) Research Program was first initiated by the European Value System Study Group which has developed into a well- established network of social and political scientists who aimed at high standards in creating and processing data.
There were four waves of study which was conducted from 1981 to 2008 in Europe and other countries to compare similarities, differences and value changes among the citizens covered by the EVS member countries. The first wave was conducted in 1981 with the participation of the European Members States of that time and the United States and Canada, in total of 16 countries/ religions in which the national datasets were combined into an international dataset.
The second wave was repeated in 1990 with the participation of 27 European countries, the United States and Canada, in which to examine the dynamics of value change in a large part of questions was repeated. In 1999, the third wave of the study was conducted in 33 European countries and religions in which a new questionaire including new issues such as solidarity, social capital, democracy and work ethics was examined based on previous experience Lastly, in 2008, the fourth wave was conducted to cover 47 European countries and it focused on a broad range of values. This research uses the data from the 2008 wave
The topics covered by the survey are as followings..First is the perception of life regarding the importance of work, family, friends and acquantainces, leisure time, politics and religion; frequency of political discussions with friends; happiness; selfassessment of own health; feelings of: excitement or interest, restlessness, pride because of compliments, loneliness, joy at the completion of a thing, boredom, feeling good, depressed or unhappy, managed everything, sadness because of criticism; feelings of the respondent at home: relaxation, anxiety, happiness, aggression or safety.
The second topic is the leisure which is the way to spend leirsure time and whom to spend it such as alone or with partners, family, friends or with colleagues or peoaple at churches or with people at sport and culture. And it is also about the frequency of political discussions with friends or with political opinion leadership or the memberships in volunteering organizations such as social welfare services, religious or church organizations, education or cultural activities, trade unions, political parties, local political actions or human rights or environment or peace movement, professional associations, youth work, sports clubs, groups, voluntary associationsconcerned with health consumption or other groups; motives for volunteering; aversion to people with other setting; feelings of loneliness.
The third topic relating to work which is the reasons for people to live in need or the importance of selected aspects of occupational work or employment status or general work satisfaction or freedom of decision making in the job or the importance of work (work ethics, scale); important aspects of leisure time; attitude towards following instructions at work without criticism (obedience work); jobs scarce: give priority to nationals over foreigners as well as men over women in jobs, able bodied people over handicapped people and forced retirement for the elderly; satisfaction with the financial situation of the household and expected situation in a year. Regarding the work Environment, there is work orientation and aspects of job satisfaction which are importance of selected characteristics of professional work such as good pay, little pressure, job security, respectable activity, flexible working hours, ability to show initiative, a lot of vacation, meeting objectives, responsibility, interesting work, meeting own skills, nice colleagues, good career opportunities, serve society, contact with people, good physical conditions of work and weekend leisure, looking forward to the work after the weekend, pride of work, family friendly, have a say, people treated equally; perceived exploitation in the workplace; general job satisfaction (scale); satisfaction with job security; use of paid days off: look for additional salaried work, training, meeting with friends and family, additional working against boredom, voluntary work, hobbies, run own business, relaxation.
The forth topic regarding religion which includes the opinion about good and evil in every one, and about justice, freedom, peace, or the feeling of being worth risking life for own country or the life of another or current frequency of church attendance and at the age of 12; importance of religious celebration at birth, marriage, and funeral; self-assessment of religiousness, belief in God, life after death, soul, hell, heaven, sin, telepathy, reincarnation, angels, devil, resurrection from the dead; stick to religion vs. explore different traditions; problems of family life, spiritual needs and social problems of the country; assessment of the importance of religion for the future; attitude towards the role of the Church in political issues; personal God versus spirit or life force; own way of connecting with the divine; interest in the sacred or the supernatural; attitude towards the existence of one true religion; importance of God in life; experience of comfort and strength from religion and belief; moments of prayer and meditation; frequency of prayers; approval or rejection of the single 10 bids by the respondents and most people; supernatural experiences: feeling of connection with someone far away, seeing events that happened far away, felt in touch with someone dead, proximity to a powerful life force, change in the way of looking at life through a psychic experience; relationship between the parents in the youth of the respondent; connectivity of respondents with both parents; strict upbringing by parents; belief in supernatural forces; ownership of and belief in lucky charms or a talisman, reading and consideration of horoscopes; attitude towards: politicians who believe in God are unfit for public office, religious leaders should let free on how people vote, and they should not influence government, ect
The fifth topic relates to family and marriage, specifically regarding trust in family, life satisfaction, agreement of opinions with the partner and among the parents with their children in terms of: religious and moral standards, social attitudes, ideal number of children; sexual self-determination; attitude towards single mothers; homosexual couples adopting children; couples living together without getting married; duty towards society to have babies; own decision to important criteria for a successful marriage such as loyalty, high income, same social background or ethnic background, ect, accepted reasons for divorce like finance problems, betrayal of the partner, sexual abuse or dissatisfaction, argument over children, opposite viewpoints, ect
The sixth topic regarding the politics and society, specifically most important purpose of the country; political interest; political participation: signing of a petition, violence against persons or things; preference for individual freedom or social equality; self assessment on a left-right continuum; social preferences; assessment of the observance of individual human rights in the country; assessing the solvability of environmental problems, crime and unemployment at the national or international level.
The seventh topic is regarding moral attitudes which are tax cheatment, corruption, bribery or money under the table, homosexuality, abortion, divorce, euthanasia, suicide, littering, driving after consuming alcohol, paying cash to avoid taxes, casual sex, smoking in public places, speeding over the limit, sex under the legal age of consent, political assassination, avoiding fare on public transport, ect.
The eighth topic is regarding- the national identity which includes geographical group, own country, the loss of power in the world for own country, and party preference; preferred immigrant policy; opinion on terrorism and immigrants and traditions (take jobs away, undermine a cultural life, make crime problems worse, strain on welfare system, threat to society, maintain distinct customs and traditions); feeling like a stranger in own country; too many immigrants personal reasons for assistance for older people and foreginers.
The nineth topic is regarding the environment which is attitude towards the environment (scale: overpopulation, disastrous consequences from human interference with nature, human ingenuity remains earth fit to live in, the balance of nature is strong enough to cope with the impacts of modern industrial nations, humans were meant to rule over the rest of nature, an ecological catastrophe is inevitable).
The tenth topic is life experiences which is the death of own child, of father or mother, the divorce; supervising function and span of control; size of company. Regarding parnter are partner was born in the country and partcountry of birth; highest educational level; employment status of on social-security of the respondent and his partner longer than three months in the last five years; number of people working in own department; labour-union membership of the respondent or his spouse; respondent is chief wage earner with parents when the respondent was 14 years old
The topic number 11 is the respondent’s partner regarding highest educational level, employment status of the partnet, partner’s profession and occupational position, duration of unemployment and dependence on social- security of the respondent and his partner longer than three months in the last five years, scale of household incomes, living together with parents when the respondent was 14 years old.
The topic number 12 is the respondents’ parents: highest educational level of father and mother; employment status of father and mother; profession of father and mother (ISCO-88) and kind of work; number of employees (size of business); supervising function and span of control of father and mother; characterization of the parents when respondent was 14 years old (scale: liked to read books, discussed politics at home with their child liked to follow the news, had problems making ends meet and had problems replacing broken things); region the respondent lived at the age of 14, present place of residence; size of town; region; region and size of town where the interview was conducted; type of habitat, ethnic group, post-materialist index 12-item, 4-item. Sweden 1999 (additional World Values Survey variables): satisfaction with public servants; economic-aid to poor countries, living with the parents and other relatives; living in a house or apartment; full time or part time employment of chief wage earner; subjective assessment of own social class, socio-economic status of respondent.
Demography: sex; age (year of birth); born in the country of interview; country of birth; year of immigration into the country; father and mother born in the country; country of birth of father and mother; current legal marital status; living together with the partner before marriage or before the registration of partnership; living together with a partner and living with a partner before; steady relationship; married to previous partner; end of relationship; divorced; number of children; year of birth of the first child; size and composition of household; living together with the parents and other relatives; living in a house or apartment; full time or part time employment of chief wage earner; subjective assessment of own social class, socio-economic status of respondent.
The survey was conducted among the following age group. The survey in 2008 was conducted on persons aged 18 years or older who owned private households, regardless of nationality and citizenship or language. Specifically in Armenia, the interview was conducted only on persons who are 15 years or older and in Finland on people who are in age from 18 to 74 years old.
Regarding the research areas, the study was conducted in countries which are as follows: Albania, Armenia, Austria, Azerbaijan, Belarus, Belgium, Bosnia-Herzegovina, Bulgaria, Canada, Cyprus, Northern Cyprus, Croatia, Czech Republic, Denmark, Estonia, Finland, France, Georgia, Germany (East/West), Great Britain, Greece, Hungary, Iceland, Ireland, Northern Ireland, Italy, Kosovo, Latvia, Lithuania, Luxembourg, Malta, Republic of Macedonia, Republic of Moldova, Republic of Montenegro, The Netherlands, Norway, Poland, Portugal, Romania, Russian Federation, Serbia, Slovak Republic, Slovenia, Spain, Switzerland, Sweden, Turkey, Ukraine, USA.
2. Methodology and processing
Regarding selection method
In the European Value Survey which was conducted in 2008, Representative multi-stage or stratified random sample of the adult population of the country who are 18 years old and older (except Armenia 15+ and Finland 18 to 74 years) were selected. Respondents were supposed to have enough command of one of the respective national language(s) to answer the questionnaire. The net sample size (in the sense of completed interviews) is 1500 respondents per country, except Northern Cyprus and Northern Ireland (with 500 interviews each), Iceland (808), Cyprus (1000), Ireland (1013), Norway (1090), Finland (1134), Sweden (1187), Switzerland (1272), France (random sample: 1501, two additional quota samples: 1570), Germany (disproportional sample East: 1004, West: 1071).
Mode of data Collection
In the EVS 2008: Face-to-face interviews with standardized questionnaire was conducted in the appropriate national language(s). In each and every country, fieldwork was conducted based on detailed and uniform instructions prepared by the EVS advisory groups. As far as the data capture is concerned, CAPI or PAPI was used in nearly all countries. Exceptions are Finland (internet panel) and Sweden (postal survey). The English Master Questionnaire was translated into other languages by means of the questionnaire translation system WebTrans, a web-based translation platform designed by Gallup Europe. The whole translation process was closely monitored and quasi-automated documented
Structure of the data file
The variable is structured according to naming conventions, missing values, administrative variables, and thematic categories
Naming conventions and missing values
To name the variables, several rules are applied according to the content and type of variables with the format as follows:
For the variables which belong to a specific thematic category were placed at the end of the category.For those variables which needs to be fit into a thematic category at a particular position were named as the previous variable plus a continuous suffix, which must be either a number or a character number unattached to the previous variable. For a variable which has a clear connection to the previous variable, the naming and labeling must be in accordance with the scheme of the EVS/ WVS dictionary which edited and standardised for each of the single wave as well as for the Longitudianl Data File in terms of providing keywords on the questions of the field questionaires. In addition, for those variables whose labels of each of the single wave compose of the respective question number referred to the questionaire. A supplement is used for constructed or country- specific variables.
The collection of missing value codes with negative values are applied throughout all over the variables.
The missing values are as follows:
– 2 No answer
– 3 Not applicable (missing values which result from filter questions; missing values in administrative variables)
– 4 Question not asked (i.e. question no questionnaire or not provided for by National Program Director)
– 5 Other missing (including system missing, wild codes, illogic answer-patterns in e.g. filters)
Structural Equation Modeling
Structural Equation Modeling is referred to a collection of unrelated computer algorithms and statistical methods which fit networks of constructs to data. The mehod are often used to assess unobservable ” latent” variables which can be measured by observed variables and a structral model that define the relationships between latent variables themselves by regression equations or more developed approaches.
This method is often used in social sciences because of the ability to construct relationships between unobserved latent variables from observable ones. For example, human intelligence can not be measured directly as height or weight, but it can be measured by items or questions according to the theory developed by psychologists. And then the method of Structural Equation Modelling is applied to test the theory using data gathered from people who took the test because in this method, ” intelligence” is the unobservable latent variable and the questions designed to measure intelligence are the observed variables.
The method also provides the numerical estimates for each of the parameters or arrows in the model to show the strength of the relationships. This method has one disadvantage compared to regression equation method in terms of the lack of widely accepted goodness of fit statistics and most of the software using Structural Equantion Modelling lack of error analysis
Moving to specify the model, there are two main components of the model which are the structural model that indicates the potential causal dependencies between the endogenous and exogenous variables and also the measurement model that shows the relations between latent variables and their indicators.
To define the latent variables, the indicators are used. The indicators of social trust are statements about people: 1. Most people can be trusted 2. Most of the time people try to be helpful 3. Most people would try to be fair. Secondly, the indicators for community involvement are the answers to the questions: do you belong to: welfare organization, cultural activities, local community action, environment, ecology, animal rights, youth work, womens group, peace movement, voluntary health organizations, Thirdly, the indicators for immigration aversion are statements about immigration, they are immigrants take away jobs from nationality, immigrants undermine country’s cultural life, immigrants increase crime problems, immigrants are a strain on welfare system, immigrants will become a threat to society or immigrants living in your country, there are too many.
To determine social trust, there are also other control variables, specifically, the determinants for social trust are: monthly household income, immigration aversion, community involvement, and control variables which are life satisfaction, age, town size, employment status. The determinants of community involvement are again household monthly income, immigration aversion, social trust, and controls variables which are importance of friends/politics, interest in politics, education, stable life location, town size. Determinants of household incomes are social trust, community involvement, and controls variables which are employment status, education, gender and town size. The variables in the measurement part and structural part as well as their abbreviations are summarized in the table below:
Latent variable Questions Abbreviation
Trust most people can be trusted _Ia165_1
most of the time people try to be helpful or mostly looking out for themselves a168_01
most people try to take advantage of you or try to be fair a168a
Civicness do you belong to: welfare orgarnization a064
do you belong to: cultural activities a066
do you belong to: local community action a069
do you belong to: environment, ecology, animal rights a071
do you belong to: youth work a073
do you belong to: women groups a075
do you belong to: peace movement a076
do you belong to: voluntary health organizations a077
Immigrants aversion immigrants take away jobs from nationality g038_av
immigrants undermine country’s cultural life g039_av
immigrants increase crime problems g040_av
immigrants are a strain on welfare system g041_av
immigrants will become a threat to society g042_av
immigrants living in your country: there are too many g045_av
monthly household income (eur 1000 ppp) x047d_ln
female respondent _Ix001_2
years of education x025_ln
employed (inactive reference group) _Ix028_r_1
unemployed (inactive reference group) _Ix028_r_2
how satisfied are you with life a170
To specify the pathways in the model, there are two kinds of relationshipes, the first one is free pathways in which the hypothesized causal relationships are tested and the second one is the established and estimated relationships based on previous studies, which are ” fixed” in the model.
According to Sewell Wright (1920s), path diagram is the pictorial representations of associations, in which causal associations represented by unidirectional arrows extended from independent variables to the dependent variables. And the causal relationships are linear and residual errors point at the endogenous variables which are determined by the variables within the model. For a proper path diagram, the total association which is simple the correlation between the two variables x and y equals the sum of the compound paths connecting them. The Wright’s rule for a compound path is that there are no loops and there are no forward and then backward arrow, and there is a maximum of one curved arrow per path.
In path analysis, there are three types of effects: direct, indirect and total effects. The direct effects are those with the association of one variable with another net of the indirect paths specified in the model. The indirect effects are those with the association of one variable with another mediated through other variables in the model, and it is computed as the product of paths linking variables. And then the total effect is simply the sum of direct and the indirect effects.
To assess if the model proposed is the best among the collection of possible models, theoretically plausible models must be specified. In order to build the model, it is necessary to take into account the theoretical reasons and also the number of data points and the number of parameters which must be estimated. An identified model is the one in which the parameters must be specified and unique and there are no other equivalent formulation given by a different parameter value. A data point is an observed variable and the parameter is the value which might be the regression coefficient between the exogenous and the endogenous variable or the factor loading. If the number of data points are less than the number of the estimated parameters, then the model is ” unidentified” because there are two few reference points to account for all the variance in the model.
To estimate the parameters, it is neccessary to compare the actual covariance matrices which represented the relationshiphs between variables and the estimated one of the best fitting model by maximum likelihood estimation or weighted least squares or asymptotically distribution free methods. After estimating the model, the next step is to interpret the model by tabulating or presenting graphically the estimated paths as a path model
It is necessary to assess the fitness of the estimated model to examine how well the model fit the data to accept or reject the model or to accept one model over the competing another. SEM programs produce output which include matrices of the estimated relationships between variables in the model and in order to assess the fitness, it is necessary to compare the similarity between the predicted data and the matrices that contain the relationships in the actual data. To do this, formal statistical tests and fit indices have been developed, among the most commonly used measure of fit are the Chi- Square which is a function of the sample size and the difference between the observed covariance matrix and the model covariance matrix. Another goodness of fit tests is Root Mean Square Error of Approximation (RMSEA) which is a population index, and no specification of of baseline model is needed, It tests a null hypothesis of poor fit If the index is smaller than 0.10, it means the model estimated is good, if the index is smaller than 0.05, that means the model is very good. Another test is the Standardized Root Mean Residual (SRMR) which is squared root of the mean of the squared standardized residuals. If the SRMR is equal to 0, it indicates ” perfect” fit, if it is smaller than 0.05, it indicate “good” fit, and if it is smaller than 0.08, it shows an adequate fit.
For those goodness of fit tests which compare the given model with an alternative model, there are Comparative Fit Index (CFI), which compares the existing model fit with a null model assuming uncorrelated variables in the model. The index helps to explain the percentage of the covariation in the data which can be explained by a given model. If the CFI ranges from 0 to 1, with 1 indicating a very good fit, and if CFI is greater than 0.9, it shows an acceptable fit. Then comes the Tucker- Lewis index (TLI) or Non- Normed Fit Index (NNFI), which is relatively independent of the sample size and if the NNFI is greater than 0.95, it indicates a good model fit, on the other hand, if the index is smaller than 0.9, it shows a poor fit.
Stata is a software package which was created in 1985 by StataCorp and serves multiple statistical research purpose such as in economics, sociology, political, science, biomedicine and epidemiology. Besides, it has a number of capabilities which are data management, statistical analysis, graphics, simulations, regression and custom programming.
The software has a command- line intergace, which enables replicating analysis. The first Stata version of 8.0 and onwards have graphical user interface which has menus and dialog boxes for users to access most of built- in commands. This enables generating code which is always displayed and helping the transition to the command line interface and more flexible scripting language.
Stata is only able to open a single data set each time which is stored random – access or virtual memory, with the limit capability of storing extremely large datasets. This is due to some extent to the internal storage ability. The format of the dataset is always rectangular, which means all the variables must have the same number of observations, or in other words, all the vectors must have the same length despite the missing values in some entries.
The software Stata can also import data in various formats such as ASCII data format ( CSV or databank formats) or spreadsheet formats such as Excel formats. Because the file formats of Stata are platform independent, users using different operating system can easily exchange datasets and programs. In addition, every version of Stata can read all older dataset formats, and not the vice versal.
Stata allows user- written commands and its extensibility can be associated with open- source packages such as software verification, technical support and professional documentation which are commercial packages.