The Political Participation and Voting Behavior of the University Student: A Statistical Approach


Shamsunnahar Tania*

Abstract: The purpose of this research is to find the status of the university student in the political activities and address the key determinants of the voting behavior. A sample of 408 students’ from Jahangirnagar University, Bangladesh shows that 46% students is registered voter; of them 77% are male. The survey indicates that 32% of female and 54% of male are voter. Fifty six percentage students think that they will probably vote in the next election whereas a significant portion (46%) is not interested to vote. The logistic regression analysis is applied to identify the most significant character that may influence voting willingness. The analysis reveals that Gender, Socio-economic status of the students, previous vote giving status, voting values and optimism or Pessimism are the significant determinants to measure the youths’ voting enthusiasm.

Keywords: Voting behavior, Enthusiasm, Values of voting, Optimism, Pessimism, Student Politics.


Voting behavior is supposed to be affected from the psychological attributes of the individual voter, besides the individual voter’s social and cultural environment. Researchers who consider the psychological perspectives believe that the main determinants of voter behavior are voter’s personal characteristics and his/her value system. Voters instinctively make choices under psychological forces like fear, aggressiveness and selfishness. Researchers including Political Science and Population Studies have done much to further understanding of the factors that influence voters to vote. Values and attitudes related to religious beliefs (Layman, 1997) and views on race are predictors (Abramowitz, 1994) of who turns out to vote in presidential elections, as are other factors such as reading the newspaper (Entman, 1989), voter knowledge (Bennett, 1994), and attending regular church services (Layman, 1997). Census data show that socio-economic status, gender, home ownership, racial and ethnic background, and age are predictors of voting in a presidential election (Jamieson, Shin, & Day, 2002).

Young generation especially university students are one of the most important parts of any election. The university students are interested in politics and are aware of the voting requirements. They collect and share information and views why and how they should vote. They play a vital role in any election. So it is important to analyze the voting behavior of the students at university level. This study aims to investigate the key characters of the students that may influence the willingness of the students to vote in the local and national election.

In this article, we have reviewed of some of the recent literature on political participation of youth. Subsequently we have presented the main dependent and independent variables that will be included in our multivariate analysis. Further, we have shown the method of analysis that is called for in this kind of research, i.e., a binary logistic regression analysis. In the next section we have described the results of our analysis, and finally, we have made conclusion with some reflections on the theoretical relevance of our findings.

Review of Literature

In this section a brief review of literature on different aspects of voting behavior has been made. Baogang (2005) found that the level of participation was influenced by three major factors: the perceived worth of the election itself, regularity of electoral procedures, and the fairness of electoral procedures. In addition to these, many other aggregate factors (such as gender, race and class as well as generational effects) are influential. The traditional individual characteristics like race, class and gender have been found to be moderate indicators of potential voter participation (Verba and Nie, 1972).

However, explanatory behavioral characteristics responsible for voter turnout described in the literature are usually centered on custom and habit formation (Yalch, 1976; Erikson, 1981; Green 2000). This method of analysis usually yields the following result: an individual’s casting a ballot in a previous election increases the likelihood that he or she will participate in the future (Green and Shachar, 2000). Assuming the habit of voting “takes root” (i.e. becomes more familiar and less daunting), an individual’s future actions (trips to the ballot box) become much easier to predict (Yalch, 1976). As a variable, race matters when examining voter turnout (Verba and Nie, 1972; Abramson and Claggett,1984). African-Americans, for example, tend to have a substantially lower voter turnout than whites across regions and election types (Plutzer and Wiefek, 2006; Mangum, 2003). Reasons for race being so highly correlated with voter turnout might include general trust or distrust in government by various minority groups, political engagement, and/or political efficacy, or any number of other factors (Mangum, 2003).

Another primary individual characteristic often used to explain voter turnout is socio-economic level. This identifier has been used in a number of ways; including as a measure of class consciousness (a pseudo-psychological variable in and of itself) manifest in times of inter-class conflict (Winders, 1999) and as a model of likelihood of voting across populations with fluctuating incomes (Filer, Kenny and Morton, 1993). Across the studies, the general consensus is that (for a variety of reasons) there is a positive relationship between income and voter participation (Sklar, 2000; Malchow, 1998). Gender, socio-economic status, civic education, the general participation climate in school, the presence of an open classroom climate and socio-economic status of the pupil significantly related to the reported willingness to vote of young people (Hooghe, Kavadias & Reeskens 2006). Dean Ladin (2007) shows that when a college-aged person has civic education, that person is more likely to be registered and vote.


Target Population & Sample

The target population of the study was the university level students in Bangladesh. This study have used data from a simple random sample of students from Jahangirnagar University. Jahangirnagar University is chosen for the convenience of the researcher. Jahangirnagar University is a public university in Bangladesh, located in Savar, Dhaka. It is the only one fully residential university in Bangladesh. The university was established in 1970 and formally launched on 12 January, 1970. The university stands on the west side of the Asian Highway, popularly known as the Dhaka-Aricha Road and spread over a land area of about 700 acres (2.8 km²). At present, the university has a total of 12,500 students, 672 teachers, 206 staffs and 1,200 other employees and there are 34 departments under six faculties and 2 institutions.

The sampling frame for this study included all the students currently studying in the university. From a pilot survey, it has been found that around 55% students (p = 0.55) are registered voter (Minimum age to be voter is 18 in Bangladesh), considering 5% alpha level (z0.025 = 1.96) and 4.5% acceptable error margin (d = 0.045), the desired sample size has been estimated by Cochran’s formula (n = z2p(1-p)/d2) as 470. So, a random sample of 470 students was selected from the database maintained by the University Registrar Office. Of the 470 students, 408 were collected completely. Rest 62 questionnaire has been found incomplete or misreported so omitted from the sample. Therefore the final sample size for the study remained 408.

The questionnaire was divided into two parts. First part covered the demographic variables included age, sex, religion, parents education and occupation etc. and university status (class level). Second part of the questionnaire contained information of political involvement of the respondent and views. In this section, participants were asked whether or not they were registered to vote and if not, whether they want to be registered. Previous voting experience was measured by each participant marking every type of election in which he or she had previously voted. Participants were also asked if they had ever participated or organized political rallies or political campaigns. Attitudes toward voting were measured in several ways. Participants were asked the likelihood of voting in the next general election on a scale of “willingness to vote,” and “unwillingness to vote.”

Dependent Variable

Students’ willingness to vote in the upcoming general election is the variable of interest in this study. It is measured by the question “Are going to vote in the next general election?” An answer with two categories is ordered to express the voting desire.

Table 1: Will you vote in the next general election?




Unwillingness to vote


Willingness to vote

Independent Variables

Independent variables as predictors of voting behavior included demographic variables and political participation related variables as provided in Table 2.

Table 2.—Variable Names and Definitions of Variables Used in the Analysis

Variable Description
Age Student’ age
Gender 0 = Female, 1 = Male
Ses (Socio-Economic Status) 1= Poor, 2=Lower Middle Class, 3=Middle Class 4=Above Upper Middle class
Religion 1=Muslim, 2=Others
Voted Ever 0 = No, 1 = Yes
Voting Values 0 = Unimportant, 1 = Important
Expect New Leadership 0 = No, 1 = Yes
Optimism Pessimism 0=Pessimistic, 1=Optimistic

Statistical Technique

Logistic regression analysis is used to identify the relationships between dependent variable and independent variables. It is used to predict a categorical (usually dichotomous) variable from a set of predictor variables. With a categorical dependent variable, discriminant function analysis is usually employed if all of the predictors are continuous and nicely distributed; logit analysis is usually employed if all of the predictors are categorical; and logistic regression is often chosen if the predictor variables are a mix of continuous and categorical variables and/or if they are not nicely distributed (logistic regression makes no assumptions about the distributions of the predictor variables). For a logistic regression, the predicted dependent variable is a function of the probability that a particular subject will be in one of the categories given a set of scores on the predictor variables.

The logistic regression model is shown as the following form:

ln(p/1-p) = βoiXi

where p = the probability of willingness to vote; (p/1-p) = odds of willingness to vote; βo = constant; Xi = vector of independent variables; βi = parameter estimate for the ith independent variable.

Binary Logistic Regression (BLR) procedures can utilize standard regression techniques to select variables (Hosmer & Lemeshow, 2000; Tabatchnick & Fidell, 2007). These include stepwise selection, in which statistical procedures are used to select variables that make the largest contribution to prediction of the outcome variable, or forced entry, in which the researcher determines which variables are included based on theory or practice. A forced entry procedure is used in this study. In BLR it is also important that no two independent variables are heavily correlated (referred to as multicollinearity). There is no definitive test for this when using categorical variables, thus conceptually similar variables were identified and one was eliminated based on which one had the smallest amount of missing data (Hosmer & Lemeshow, 2000; Tabatchnick & Fidell, 2007). Continuous variables can be included, but require checking the assumption of “linearity in the logit,” or checking the relationship between the continuous variable and the outcome variable (Hosmer & Lemeshow, 2000). Interaction terms should also be explored to determine if the outcome varies by levels of an independent variable, however, this requires somewhat advanced techniques that are explained in Hosmer and Lemeshow (2000) and Tabatchnick and Fidell (2007) and are not covered here.

Data Analysis and Results

Demographic Characteristics:

From all the respondents, about 67% are male and 33% are female. About 85% of the respondents are Muslim and rests of them are from other religious group; 57.8 percent are reported under age 20, 25.5 percent are in the age group of 20-22 and rest 16.7 % are above 22 years. Of the student 5.9% have reported that their father are illiterate, only 51% of the student’s father are well educated with academic level graduate or above; in case of mother education level 6.1% told that their mother are illiterate, 18.4% have graduate or above educated mother.

Figure 1: Gender & Socio-Economic Status of the Students (Source: Field Survey)

Political Participation

The Political Participation

Forty six percent of the respondents were found registered voter; however, 32% of the female students were registered as voter whereas about 54% of the male were voter. Of the registered voter 77% were male and only 23% were female. Of the registered voter only 12% have voted in any type of election. In response to the likelihood of participants voting in the next general election 56 percent think they probably will vote; others declared that they probably will not vote.

Figure 2: Registered to Vote and Willingness to Vote (Source: Field Survey)

to vote

Only 21% of the students were involved in politics actively, 23.8% have attended various political campaign and 30.6% have attended political rally. When we asked them about their values regarding vote 64% of them replied that voting was important while others thought it unimportant.

Table 3: Gender Vs Registered to Vote





Registered to vote







Source: Field Survey

In order to gain further insight on the voting behavior of the students, they were asked whether or not they believed that voting makes a difference. The vast majority (78%) believed that their vote surely will make a difference. Our survey investigated that most of the student (76.2%) regarded voting as their right and responsibility; 10.3 percent expressed that they did vote for ‘To get those I trust elected’ 9.6% reported voting as their sacred duty (Table 4).

Table 4: Why do you vote?




Right and responsibility



To get those I trust elected



My sacred duty



Just following others



at the request of cadres



For other reasons



Source: Field Survey

Very interestingly, student showed sex preference of the candidate 55.1% of the total surveyed students declared that they want a male candidate and 10.8% preferred female candidate 34.1% have no option; they are willing to accept either gender. 

Table 5a: Preference of Gender of the candidate







No option


Table 5b: Preference of New candidate






Source: Field Survey

Very important experience on whether the respondent will accept a new comer as their leader; they responded positively towards new comer; 88% have mentality to accept new face but 12% think old politician will do better.

Correlation of Political Participation with Demographic Character

In bivariate analysis part, cross tabulation technique has been applied to identify whether there exist any significant difference in sex of the respondent with other concomitant variables. The results of the bivariate analysis are presented in table-7 of this section.

Table 7: Association of various political participations with gender of the respondent


Value of Chi-Square



Vs Registered to vote




Vs Active participation in politics




Vs Importance of voting




Vs  attending electoral campaign




Vs attending political rally




Vs Preference of fresh candidate




Vs Likelihood to vote




Source: Field Survey

The table shows that gender of the student has a remarkable influence on voter registration. Significant difference also exists in political involvement on sex of the respondent. We have found significant correlation between likelihood to vote and gender of the student. The preference sex of the candidate, perhaps one of the most interesting findings of this research, is strongly related with sex of the respondent. The differences between male and female opinions of valuing the voting show no significant differences. Comparisons between the two groups show similar responses in case of preference of new & fresh candidate.

Logistic Regression Analysis

In this study, we have tried to investigate the key determinants that influence the willingness to vote in the next general election. We have examined the variables discussed in the table 2 as predictors. We have used Binary Logistic Regression (BLR) analysis which comprises few steps. Firstly, we test the overall relationship of the predictors. Secondly, strength of BLR relationship has been established. Thirdly, we have evaluated the usefulness of logistic model and finally significant relationship between the independent and dependent variables has been checked and interpreted.

The overall test of relationship between the dependent and independent variables of the final model reveals that the probability of the model chi-square (137.488) was 0.000, less than the level of significance of 0.05 (i.e. p<0.05). Therefore, the null hypothesis that there is no difference between the model without independent variables and the model with independent variables was rejected.

Once the relationship is established, the next important thing to do is to establish the strength of binary logistic regression relationship. In this case, Cox & Snell R Square and the Nagelkerke R square value, provide an indication of the amount of variation in the dependent variable. These are described as pseudo R square. The distribution reveals that the values are 0.289 and 0.387 respectively, suggesting that between 28.9% percent and 38.7% percent of the variability is explained by this set of variables used in the model. The classification accuracy rate is 77% which suggests that the model is useful.

The likelihood ratio test evaluates the overall relationship between an independent variable and dependent variable. While, the Wald test evaluates whether or not the independent variable is statistically significant in differentiating between two groups in each of embedded binary logistic comparisons.

Table 8 presents the parameter estimates for the final model. The Wald Chi-Square statistic tests the unique contribution of each predictor, in the context of the other predictors — that is, holding constant the other predictors. We see that Gender, Socio Economic Status, Voted Ever and Voting Values are the significant variable to predict voting behavior with the willingness to vote in the next election.

The odds ratios less than 1 indicate a lower likelihood for the event of interest; odds ratios greater than 1 indicate greater likelihood for the event of interest. Male Students were 2.45 times as (more) likely to have willingness to vote compared to female students. Students with different age are indifferent to or not to vote. Upper middle class or rich students are equally likely to vote compare to poor students; however, lower middle class and middle class students are 4.25 and 3.92 times more likely to exercise voting power than poor students. The students who have already applied their voting power in any format of elections are 3.11 times more likely to vote compare to the student who did not vote yet. Very usually, the students who think that vote is important are much more interested to vote which is about 10.71 times more compare to the students who think voting unimportant. Surprisingly, the students who expect new leadership and who do not are equally likely to apply voting right.  The optimistic students are 2.89 times more likely to have willingness to vote in the next general election.

Table 8: Parameter Estimation of BLR of willingness to vote Vs unwillingness















  Female (Ref)
















  Others (Ref)

Socio Economic Status*

  Lower Middle Class







  Middle Class







  Above Upper Middle Class







  Poor (Ref)

Voted Ever*








  No (Ref)

Voting values*








  Unimportant (Ref)

Expect New Leadership








  No (Ref)

Optimism vs Pessimism*








  Pessimistic (Ref)








OR=Odds Ratio, *Significant Variables (p<.05)

Concluding Remarks

The university students are the most conscious segment of a nation. Therefore, the behavior of the student in the voting system is important. We have seen, 46% of the students are registered voter. Male registered voter are more than triple than female. In the voting system female representation is very low; only 32% of the female students are registered as voter whereas about 54% of the male are voter. The willingness to vote is not satisfactory; about 46% of the students are not interested to vote. Only 21% of the students are actively involved in politics. The students’ views of the voting are very important. About 64 percent stated that it was civic responsibilities. Very interestingly, student showed sex preference of the candidate; 55.1% of the total surveyed students opined that they preferred a male candidate on the other way 10.8% preferred female candidate while 34.1% did not have any option. Very important experience on whether the respondent will accept a new comer as their leader; they responded positively towards new comer; 88% had inclination on new face on the other hand 12% suggested for experienced politician. We have found that the political participation was very much biased on sex issue. Gender of the students had been identified as significant predictor for voting behavior.

In this study we have analyzed the impact of various demographic variables on the willingness to vote of the students at university level. Our analysis confirmed most of the existing literature. Sex of the student, Socio Economic Status, Voted Ever and Voting Values, has a strong and significant effect on the willingness of young people to participate in the voting process. Age, Religion, Expectation of new leadership and Optimism did not have a significant impact on voting behavior of the university students.


Abramowitz, A I. (1994). Issue Evolution Reconsidered: Racial Attitudes and Partisanship in the U.S. Electorate. American Journal of Political Science, 38 (1), 1-24.

Abramson, P. R. and Claggett, W. (1984): Race-Related Differences in Self-Reported and Validated Turnout, The Journal of Politics, Vol. 46, No. 3, pp. 719-738.

Baogang HE , A survey of Voting Behavior and Political Participation in Zhejiang, Japanese Journal of Political Science, 7(3) 225-250,  Cambridge University Press.

Dean Ladin (2007), Civic Education and the Voting Behavior of College Students, Final Paper under Professor Kent Portney.

Entman, R. M. (1989). How the Media Affect What People Think: An Information Processing Approach. The Journal of Politics, 51 (2), 347-370.

Erikson, R. S. (1981): Why do People Vote? Because they are Registered, American Politics Quarterly, Vol. 9, pp. 259-276.

Filer, J.E., Kenny, L. W. and Morton, R.B. (1993): Redistribution, Income, and Voting, American Journal of Political Science, Vol.37, No.1, pp.63-87.

Green, D. P., Shachar, R. (2000): Habit Formation and Political Behavior: Evidence of Consuetude in Voter Turnout, British Journal of Political Science, Vol. 30, No.4, pp. 561-573.

Hooghe M, Kavadias D & Reeskens T (2006): Intention to Vote and Future Voting Behavior. A Multi-Level Analysis of Adolescents in 27 Countries. Annual Meeting of the American Political Science Association, Philadelphia, August 31-September 3, 2006

Hosmer, D. W., & Lemeshow, S. (2000). Applied logistic regression (2nd ed.). New York: John Wiley & Sons.

Jamieson, A., Shin, H. B, and Day, J. (2002). Voting and Registration in the Election of November 2000: Population Characteristics. U. S. Census Bureau Report.

Layman, G.C. (1997). Religion and Political Behavior in the United States: The Impact of Beliefs, Affiliations, and Commitment From 1980 to 1994. The Public Opinion Quarterly, 61 (2), 288-316.

Malchow, H. (1998): Predicting Voter Turnout: A New Targeting Approach, Campaigns & Elections, July.

Mangum, M. (2003): Psychological Involvement and Black Voter Turnout, Political Research Quarterly, Vol. 56, No. 1, pp. 41-48.

Plutzer, E. and Wiefek, N. (2006): Family Transitions, Economic Status, and Voter Turnout Among African-American Inner-City Women, Social Science Quarterly, Vol. 87, No. 3, pp. 658-678.

Sklar, H. (2000): Booming Economic Inequality, Falling Voter Turnout, Z Magazine.

Tabatchnick, B. G.,& Fidell, L. S. (2007). Using multivariate statistics (5th ed., pp. 481–498). Boston: Pearson Education, Inc.

Verba, S. and Nie, N. H. (1972): Participation in America: Political Democracy and Social Equality. New York: Harper and Row.

Winders, B. (1999): The Roller Coaster of Class Conflict: Class Segments, Mass Mobilization, and Voter Turnout in the U.S., 1840-1996, Social Forces, Vol. 77, No. 3, pp. 33-862.

Yalch, R. F. (1976): Pre-Election Interview Effects on Voter Turnout, Public Opinion Quarterly, Vol. 40, No. 3, pp. 331-336.

*   Assistant Professor, Department of Business Administration (Marketing), Stamford University, Dhaka, Bangladesh

Leave a Reply

Your email address will not be published. Required fields are marked *