Students’ Attitude Towards Reading and Its Associated Factors in Samara University, North East Ethiopia

: Reading is mentally stimulated, attain knowledge and ideas, reduce stress in their lives, advance their vocabularies and memory, and develop keener analytical skills. All of these skills and abilities are critical to a student's success in school and life. Accordingly, this study was intended to identify and analyze the factor that affect student’s attitude towards reading at Samara University. This study used cross sectional study design and 345 students were nominated by via stratified random sampling method and the data have been collected using structured questionnaires. Binary logit model was used under this study. Descriptive statistics result out 345 students 55 (15.9%) had negative attitude towards reading and the remaining 290 (84.1%) had positive attitude towards reading, most students had a positive attitude towards reading. Binary logit regression result that showed: educational level of student’s family, palace of coming students, participation of group discussions and students reading schedule are significant and significantly affect attitude toward reading. This study suggested that: students should have a reading schedule, students must be attending in class & participate group discussion, and the university should provide materials that control the effect of weather condition such as air cooling (AC) and ventilator.


Introduction
Reading is the identification of the symbols and the association of suitable meaning with them. It requires identification and comprehension. Understanding skills help the learner to understand the meaning of words in isolation and in context [8]. He trusts reading is a process of thinking, evaluating, judging, imagining, reasoning and problem solving [8].
Reading plays a very important role in our lives. It is so much a part of everyday living that one can hardly imagine a life without it. In the age of the Internet and information technologies, reading retains its importance as an essential skill for learners of any language [3]. Reading is amongst the most vital determinants in developing an individual's vision that shapes his or her personality and that makes him or her become closer to other individuals. Reading makes individuals truly free and protects them from unawareness and false beliefs [9]. Also, skills in reading enable individuals to benefit from educational activities, and to participate fully in the social and economic activities in which they take part [8].
Reading has been described in a variety of ways to explain the process of what happens when one reads and how one comprehends a text. However, a widely accepted explanation of reading is not found in the literature [6].
Grabe and Stoller [6] define reading as the ability to draw meaning from the printed page and interpret the information appropriately. Reading is a psycholinguistic process in that it starts with a linguistic surface representation encoded by a writer and ends with meaning which the reader constructs [5]. The common characteristics of definitions relating to reading are that it is an activity that is realize through comprehension. Reading is not an activity that takes place individually but one that encompasses many events and situations [9].
Reading is a crucial tool of learning. Every course of study is talented partly through reading. Lecturers in the University usually have high expectations of a student's" ability to cope with the demands of reading. However, reading as a practice and an art has tended to diminish. The general expectations of these students reading ability have not been encountered. Students have a key role to play for them to succeed academically. Required information, skill and more information can be acquired through reading [2].

Operational Definitions
Attitude -a mental position, a feeling or emotion towards the acts of reading.
Reading understanding -the ability to comprehend and interpret a written text.
Reading -the process of constructing meaning from writing This is the tricky of most students have that contributes to their poor routine in reading attitude or lack of proper reading habits. For a first-rate performance, there is the need for the student to form good reading and study habits. At present, due to the influence of the mass media, people do not show much interest in reading books; magazines and journals, among others [7].

Objective
The aim of this research was to determine the reading attitudes of samara university students based on some variables. The main objective of this study was to determine the factor that affects students' attitudes in the direction of reading at samara university students. In accordance to that, a) To determine the significant factor that affects students' attitude towards reading using logistic regression. b) To determine whether a significant difference in reading attitude among male & female undergraduates.

The Study Area
This study was conducted in Samara University. Samara University (SU), Ethiopia, is a government higher education institution located in the small town of Samara (population range of 2,500-9,999 inhabitants), Afar Region. Officially attributed by the Ministry of Education (MOE), Ethiopia, Samara University is one of higher education institution. It offers courses and programs important to officially recognized higher education degrees such as bachelor's degrees in several areas of study. Since its beginning in 2008, Samara University is making great jumps towards producing competent and dynamic graduates who fulfill the needs and aspirations of the people. SU also provides several academic and non-academic facilities and services to students including a library, as well as administrative services.
Samara town is the hottest in Ethiopia that is placed 588km far from the northeast of Addis Ababa. It is considered by an arid and semi-arid climate with low and erratic rainfall.

Sampling Strategy
The sampling design used in this study was cross-sectional data by self-administered questionnaire for the purpose of performing this research in the year 2021 G.C. The target population of the study was Samara University three colleges purposely selected, Collage of natural and computational science, Collage of Engineering and Technology, & Collage of Veterinary Medicine students.

Method of Data Gathering
In order to answer the research questions in the study, the data was gathered by using a self-administrated structured questionnaire. The attitude level towards reading gomleksiz (2004) was used in the study in order to distinguish the reading attitude of the student.

Sampling Method
The sampling technique used for this study was stratified random sampling. A stratified probability sampling technique was used when separate information is needed among the study population. College of students was used as strata for the purpose of sampling. But in this study due to a shortage of time and budget, we used only three colleges selected randomly from the strata. The major reasons using stratified random sampling technique were to increase the efficiency of the sample by dividing the study population into strata to create as great homogeneity as possible within each stratum and as a marked difference (heterogeneity) as possible between the strata. This also increases the accuracy of the estimator and allocates the sample to each college. In stratified sampling, the population of size N is divided into i sub-population (strata) of size N 1 N 2 , & N 3 and the arbitrary sample of size n is drawn from stratum i; i=1,2,3.

Sample Size Determination
Due to shortage of time and other resources the whole populations were not feasible to study. As a result of the representative sample size were taken. To calculate the sample size, we used that stratified random sampling method. Since the sampling method is stratified random sample, the sample were taken from each stratum (in this case the strata are CNS, CET, CVM, students). Sample of undergraduates from each stratum was taken by using the way of proportional allocation which means that the sample allocated to each stratum was comparative to the total number of populations in the stratum.
Where: N=3413: total number of students, n=345 sample size needed for the study p=proportion of students having positive reading attitude=0.5, q=proportion of student having negative reading attitude=0.5, d=the error term=0.05 and Z α/2 =Z0.025=1.96.
Proportional allocation is used when the stratum has big difference in size. It means a small sample selected from a small stratum and large sample taken from a large stratum and the sample size in each stratum is fixed. The samples were calculated by the formula: ( = )( ) .

Dependent Variable
The students' attitude towards reading (positive, negative) was determined by gomleksiz, 2004 attitude scale towards reading.

Symbol of variable Category
Students attitude towards reading Y 1=positive attitude 0=negative attitude

Independent Variables
Independent variables are the explanatory variables that provide to explain the response variables -student's attitude towards reading.

Descriptive Statistics Analysis
This method often used for describing the characteristics of the sample and the study variables were displayed in the form of frequency distribution, percentage [1]. So, we were applying this method in this study.

Inferential Statistics
Inferential statistics is statistical method deals with making inferences or conclusions about a population based on data obtained from a limited number of observations that come from the population. Inferential statistics contains of the estimation and hypothesis testing [1]. In this study chi-square test and binary logistic regression were applied to measure the association between two categorical variables.

Binary Logistic Regression
Binary logistic regression is used when the dependent variable is dichotomous and the independent variables are either continuous or categorical variables. The assumption in binary logistic regression is that observations are independent of each other. Violations of the assumption of independence of observations may result in improper statistical inferences due to biased standard errors. The binary logistic regression function is defined as:

Odds Ratio
Logistic regressions work with odds so it is necessary to define both odds and odds ratio. The odds are simply the ratio of the probabilities for the two possible outcomes. If 0 is the probability that positive attitude towards reading, then 1 -0 is the probability that negative attitude towards reading: odds =  The ratio of the odds from the two rows, is odds ratio.
If the parameter β j is positive then the odds increase and if the parameter β j is negative, the odds decrease. Samara University, North East Ethiopia

Maximum Likelihood Estimation
The technique of maximum likelihood used to estimate the parameters from the linear transformed model ΧE′β. β be the estimate of the model parameters.
Where the matrix v is n×n diagonal matrix holding the estimated variance of each data points on the main diagonal; that is the i th diagonal element of v is v RR = E0(1 − 0) the estimated value of the linear predictor is Ti=ΧE′7, and the fitted value of the binary regression model.
The likelihood specifies how likely the observed sample is a function of parameter values. The likelihood of a set of parameter values, 7 given outcomes x is equal to the probability of those observed outcomes given these parameter values that is. The likelihood functions the joint probability density function of random variable but, it's observed as the meaning the parameter given the realized random variable.

Model Diagnostic Test
After the model is fitted the next vital step is checking the model adequacy. There are several steps elaborate in assessing the appropriateness, adequacy and usefulness of the model. First, the overall goodness of fit of the model is tested. Second, the importance of each of the explanatory variables is measured by carrying out statistical tests of the significance of the coefficients.

The Likelihood Ratio Test
The likelihood ratio test, also called the log-likelihood test, is based on -2LL (deviance). The likelihood ratio test is a test of the importance of the difference between the likelihood ratio (-2LL) for the fitted model and the likelihood ratio for a reduced model. This difference is called "model chi-square".

Test of the Overall Goodness of Fit
Test of the overall goodness of fit is used to assess the overall goodness fit of the model. likelihood ratio test aspects at the model chi-square (chi square difference) by subtracting deviance (-2LL) for the full model from deviance for the intercept-only model. The degrees of freedom in this test equal the number of terms in the model minus one. This is the same as the difference in the number of terms between the two models, since the null model has only one term.
Model chi-square measures the improvement in fit that the explanatory variables make compared to the null model. The likelihood ratio test is thus a test of the overall model. The overall test statistic for likelihood ratio test is given as: Likelihood ratio test is u B = −2ln(v wxx v y ⁄ ) = −2 ln(vv wxx vv y ) Where,v wxx is likelihood of the null model and L k is the likelihood of the model with k factors.

Test of Individual Parameters
The likelihood ratio test measures the overall logistic model but does not tell us if particular explanatory variables are more important than others. A non-important likelihood ratio test indicates no difference between the full and the reduced models. Note that the likelihood ratio test of individual parameters is a better condition than the alternate Wald test when considering which variables to remove from the logistic model.

The Wald Test
The Wald test is an alternative test which is commonly used to test the significance of individual logistic regression coefficients for each covariate variable (that is, to test the null hypothesis in logistic regression that a particular logit (effect) coefficient is zero. (Alan Agresti, 1990) The Wald test statistic is:

Hypothesis Testing
The hypothesis testing for the i th explanatory variable. H0: The coefficient of parameter is no associated with the predictor is equal to zero (7 V = •€•8) H1: The coefficient of parameter is associated with the predictor is not equal to zero (7 V ≠ •€•8) Where: -7 … is the estimated parameter for the population7 … , se (7 … ) is the standard error of 7 … .

Decision
If p-value is less than †-value=0.05 H O will be rejected at a given level of significance and we can conclude that the independent variable Xi has significant effect on the probability of success (binary response).

Analysis of Inferential Statistics
It is discussed about the model summary, in order to identify variables related with students' attitude towards reading in binary logistic model was used. Moreover, the joint impact of all predictor variables on the dependent variables also determines by using the Nagelkerke ‡ B which is described in the model summary.

Results and Discussions
In this section, the results of this study were offered in two key portions. In the first section, the results of the descriptive statistics for the sampled students were presented. The result of the binary logistic regression in the second section.

Summary of Descriptive Statistics
From the above table 4, the smallest age of students is 21years old and the highest age is 30 years old. The average age of students is 24.35 years old. The standard deviation is 1.873.  From the above table 5, among 345 students 48.4% are female and 51.3% are males. When we see place of coming 47.5% from rural students, 11% from suburban students & the remaining 41.4% are from urban students. 31.6% students at tend the class sometimes but 68.4% students are always attending in class. 22.6% of students are always attending in group discussion, 58.3% students attend sometimes and the remaining 19.1% students never attend in group discussion. About 222 (64.3%) students were coming from almost medium economic status of family and the remaining 31 (9%) and 92 (26.7%) students are coming from high and low economic status of family respectively. From our respondents 64 (27.2%) were college of natural and computational science students, 232 (67.2) were college of engineering technology students and 19 (5.5%) were college of veterinary medicine students.
When we see effect of weather condition, 70 (20.3%) of the respondents said that weather was comfortable for them to read, 14 (7.6%) were no such much & the remaining 247 (71.6%) are strict fact on reading. 164 (47.5%) students had reading schedule &181 (52.5%) had no reading schedule. 106 (30.7%) students were read sometimes, 142 (41.2%) were read usually and 97 (28.1%) were read for long time.
According to the figure 2, out of 345 students 55 (15.9%) have negative attitude towards reading and the remaining 290

Analysis of Inferential Statistics
In this section it was discoursed as the model summary, in order to identify variables related with students' attitude towards reading in binary logistic regression model was used. Moreover, the joint influence of all explanatory variables on the dependent variables also regulates by using the notion of Nagelkerke ‡ B which is explained in the model summary. Adequacy of overall model fit in binary logistic regression is the likelihood ratio test, which is simply the chi-square difference between the null and the model holding the explanatories. Under Model Summary the -2 Log Likelihood statistics is 207.074. This statistic measures how poorly the model predicts the student's attitude towards reading, the smaller the statistic the better the model. The result of Cox and Snell ‡ B indicates that 44.2% of the variation in the dependent variable is explained by the independent variable which is supposed to be good enough. Table 7. Goodness of fit test of the model.

Hosmer and Lemeshow Test
Step Chi-square text value Degrees of freedom Significance 5.122 8 .745 Source: own computation, 2021.
The "Hosmer and Lemeshow Test" is the extent of fit which estimates the goodness of fit between predicted and observed probabilities in classifying the dependent variable.
since the p-value is 0.745 which is insignificant, that is 0.745 > p-value (0.05), so the fitted binary logistic model is good fit to the data. The other way of evaluating of the goodness of the fitted model is to see how well the model classifies the observed data. In the above table 8, cases with probabilities greater than or equal to 0.50 are predicted as having the event; other cases are predicted as not having the event. The results in above table indicate that (38.2%) of the respondents which is Negative attitude towards reading were correctly predicted and (95.9%) of respondents which was positive attitude towards reading were correctly predicted. Our model shows an overall percentage (86.7%) of correct classification of respondents in their attitude towards reading either positive or negative. That is the fitted model has an overall predictive accuracy of 86.7%.

Binary Logistic Regression
Variables in the Equation for the relationship between students' attitude towards reading and covariates. From the above table 9, it is observed that the estimated odds ratio is 2.779 implies that the odds of female students is 2.779 times more likely to had positive attitude towards reading compared to male students remaining constant for other variables in the model. In this study female students had more attitudes towards reading.
The coefficient of place of coming (urban, 1.654) is positive implies that respondents from this group are more likely to had positive reading attitude than from students coming from the area suburban. The estimated odds of the palace of coming from urban is 5.230 times more likely to had positive attitude towards reading than students come from suburban (see table 9). According to Palani (2012), if students were given the appropriate training at school level, they can easily develop potential in reading. Teachers do not only teach in school. They play a vital role in inculcating positive attitudes that shape students into serious readers. If students were instilled with the values of reading, there is a very high possibility that they would continue to read for the rest of their life. In conclusion, students' attitudes towards reading affect their reading habits [9].
The odds ratio and confidence interval for student's class attending that is always is 5.268, 95% CI (2.060, 13.473). Students class attending which is always 5.268 times greater positive reading attitude, than students class attending sometimes. (See table 9). Patrick et al. (2007) found students perceptions of dimensions of their classroom social environment, including affiliation, cohesion, fairness, mutual respect, and support from teachers and students are associated consistently with adaptive motivational beliefs and achievement behaviors.
The estimated odds ratio of participating group discussion (always) is 2.873, indicates that, attending group discussion always 2.873 times more likely to have positive reading attitude compared to never participating group discussion controlling other variables in the model (see table 9). The recent emphasis placed on improving children's English Language Arts test scores often leads teachers to ignore the role of student attitudes in the process of becoming literate. In order for students to develop into effective readers, they must possess both the skill and the will to read. As noted by Wigfield and Guthrie (1997), "motivation is what activates behavior." Student attitudes toward reading are a central factor affecting reading performance [10].
Regarding father's/mother education level the estimated odds ratio 1.0622 indicates that those students whose family level of education illiterate were 1.6022 less likely to positive attitude towards reading compared to those family educational level diploma and above controlling for other variables in the model (see table 9).
According to the findings of this study the coefficient of whether condition (strict fact) is negative which can be interpreted as respondent which have an influence less likely to positive reading attitude compared to no influence of whether condition. The estimated odds of whether condition (strict fact) is 0.135 less likely positive reading attitude with the reference group (see table 9).

Conclusions and Recommendations
Reading attitude towards as an individual's feelings about reading, caused learner to approach or avoid a reading situation. A positive reading attitude helps the students in many ways. The students will not be bored in complying their reading assignments. This study was undertaken to determine the effect of socio-demographic characteristics on the attitude of students towards reading. This study tried to investigate reading attitude and its determinants factor by using a binary logistic regression model. As the result obtained more than half students have positive reading attitude and from the logistic regression analysis it was also concluded that the odds of sex of students, students place of coming, class attending, participation of group discussion, educational level of family and effect of weather condition are significant predictors on positive reading attitude in Samara university students.
Based on the above finding, this study was recommended the following implications: students should had read schedule and use this schedule. Students must be attending in class & group discussion. Families should create awareness about reading &teach for their children. The campus should prepare reading workshop, festival about reading. The campus should compliance materials that control the effect of weather condition such as air cooling (AC), ventilator.