Academic goals in blind and partially sighted students : a differential analysis and goal profiles

The present work was based on goal orientation theory that represent the reasons to perform a task and the criteria to evaluate it. However, few studies have explored goal orientations in students with visual impairment. The objective of this study was to determine whether the goals pursued by blind students different from those of partially sighted students. Participants comprised 171 students, from different stages of the education system (from Primary Education to University Education) with ages between 8 and 27 years old, of which 109 were blind, and 62 partial visual impairment, which respond to an adaptation of the Achievement Goal Tendencies Questionnaire. The results indicate that students with total visual impairment have lower average scores on social reinforcement goals, not in contrast to learning goals, achievement goals and performance goals, where no significant differences were found. Three different goal orientation profiles were obtained. The first stands out for obtaining high scores in learning goals; the subjects of the second group are defined by obtaining high scores in learning goals, social reinforcement goals and performance goals; and the subjects of the third group are characterized by their low scores on learning goal, social reinforcement goals and achievement goals. The results of the discriminant analyses show that there are differences in these profiles depending on the degree of visual disability and that the achievement goals do not have discriminating power. These findings have educational implications as regards motivation.

According to the two-dimensional model (Ames, 1992;Dweck, 1986), there are two types of goal: learning goals (focused on understanding and learning) and performance goals (focused on self and performance).Learning orientation has been considered more adaptive and has been associated with a series of positive mediators in learning; in contrast, performance orientation has been viewed as less effective and has been related to negative mediators.
Also, a four-dimensional model has been proposed that classifies goals into four categories: task goals, goals of self-assessment or self-orientation, goals of social assessment and reward goals (Núñez, González-Pienda, González-Pumariega, García & Roces, 1997).Task goals lead students to learn and develop and improve their abilities.The goals of ego orientation lead the student to seek the satisfaction experience that follows success and avoid the negative feelings that accompany failure.The goals of social assessment make the student seek the assessment and approval of others.And reward goals lead the student to obtain awards or avoid losing situations or things that value.Several investigations have identified these four kinds of goals (Valle et al., 2003a(Valle et al., , 2003b;;Valle, Cabanach, Rodríguez, Núñez & González-Pienda, 2006;Valle et al., 2008Valle et al., , 2009)).
However, this conceptual and empirical approach is not uniform but rather has evolved from normative concepts towards other, multidimensional ones.From this perspective, the goals are defined as cognitive representations of what the students see them as trying to achieve.What is important is the content of the students' goals and how they strive to achieve academic and social outcomes (Wentzel. 1993).It is argued that motivation is related to multiple goals and that the same student can pursue different goals at the same time (Wentzel, 1999(Wentzel, , 2000)), and it has been established that these give rise to different motivational profiles (Alonso-Tapia, Huertas & Ruiz, 2010; Barca-Lozano, Almeida, Porto-Rioboo, Peralbo-Uzquiano & Brenlla-Blanco, 2012;Barca, Peralbo, Porto, Marcos & Brenlla, 2011;Cecchini, González, Méndez-Giménez & Fernández-Río, 2011;Navas & Sampascual, 2008;Pekrun, Elliot & Mayer, 2009;Phan, 2012).It appears that academic goals diversify with an increasing the level of education.Thus, for example, Navas, González and Torregrosa (2002) found the three classic goals (learning, social reinforcement and achievement) in primary education students, four types of goal in secondary education students, five in postcompulsory secondary education students and up to six different types of goal in university students.
Although goal orientations constitute an important motivational factor, they have not been studied in students with visual impairment, and this was the objective of the present study.
Students with visual impairment must use their other senses to comprehend their surroundings and act with minimal delay compared to sighted people (Herrera & Ramírez, 1999;Rosa, Huertas & Blanco, 1993).Visually impaired people need the help of others to apprehend the context, at least initially.Consequently, their satisfactory development depends on adults' attitudes towards them (Navas, Jover & Castejón, 2017;Pelechano, De Miguel & Ibáñez, 1995;Warren, 1984).Social interaction is thus extremely important for people with visual impairment, because they are more dependent and have reduced possibilities of movement and of understanding their physical and social environment (Bieber-Sucht, 1991;Rosa et al., 1986;Van Hasselt, 1983;Warren, 1984).Therefore, orientation, mobility and motivation are crucial capacities for these people.
The restrictions experienced by one student with visual impairment cannot be generalised to all other visually impaired students.The most influential factor is the level of visual loss (Núñez, 1999;Verdugo & Caballo, 1995).
Students with visual impairment have unique educational needs.Imitation is limited or non-existent, and therefore learning requires greater effort and teaching activities that prioritise tactile, auditory and propioceptive perceptions (Navas et al., 2017;Núñez, 1999).
One might think that in order to prove their worth to others and to themselves, blind and partially sighted students would be more motivated to carry out their academic activities.Indeed, some studies (Holgado, Navas & Jover, 2012;Jover, 2006Jover, , 2009;;Jover, Navas & Sampascual, 2008) have found that people with visual impairment have a different goal orientation structure to that obtained for sighted students.
This raises the question of whether the degree of visual impairment leads to differences in goal orientation.Consequently, the objective of the present study was to analyse whether the goals pursued by blind students differed from those of partially sighted students, and to determine motivational profiles as a function of degree of impairment.

Participants
Participants comprised 171 students (57.3% boys and 42.7% girls) with visual impairment, selected by means of non-probability convenience sampling among affiliates of the National Organisation of Spanish Blind People (Spanish acronym: ONCE).Of these, 36.3% presented total visual impairment, with ages ranging from 9 to 26 years old (M = 15.81 ±.56; DT = 4.43) and 63.7% presented partial visual impairment, with ages ranging from 8 to 27 years old (M = 14.32 ±.36; DT = 3.80).All participants were students enrolled in different courses, according to the distribution of the Spanish educational system: 28.7 % from Primary Education (from the fourth to the sixth course), 49.1 % from Compulsory Secondary Education (from the first to the fourth course), 9 % from Bachelor (first and second courses), 10.8 % from University Education, and 2.4 % from Vocational Training.

Instruments
We administered the Achievement Goal Tendencies Questionnaire developed by Hayamizu & Weiner (1991), using the version created by González et al., (2000) which is tailored to students with visual impairment, either by using large print for partially sighted students or Braille for blind students.The questionnaire lists 20 reasons for studying, and subjects indicate the reasons why they study using a 5-point Likert type scale where 1 means "never" and 5 means "always or all the time".The questionnaire has obtained a reliability of between .89 and .71.Given the characteristics of the participants, an exploratory factor analysis (EFA, principal factor with varimax rotation) was performed.The instrument consists of four factors (Holgado et al., 2012).Table 1 shows the factors, the items that compose them, the percentage of variance they explain, and their reliability (Conbrach alpha reliability coefficient).This factorial structure was validated by applying confirmatory factor analysis (Jover, 2009), yielding the following goodness of fit indices: χ 2 (gl = 161; p = .001)= 573.85;RMSEA = .08;GFI = .98;AGFI = .97.These results indicate that the four-dimensional model obtained in the EFA fits the data reasonably well.

Procedure
Students were selected through the directors of various Educational Resource Centres (ERC) affiliated to ONCE.The questionnaire was administered by peripatetic teachers and ERC coordinators.Teachers were instructed on how the questionnaire should be completed (anonymously, in a single session, in the students' usual classroom, and expressly indicating that the students' answers would not affect their marks and that therefore they should answer as honestly as possible).At the same time, students received an assurance that their anonymity and right to privacy would be maintained.

Variables
The variables considered were: a) academic goals (learning, social reinforcement, performance and achievement), and b) students' status (total or partial visual impairment).

Design and data analysis
A basic correlational design was employed, since there was no intentional manipulation of the variables and the subjects were selected using convenience sampling, participated on a voluntary basis and gave their verbal consent.Correlational, differential and discriminant analyses were performed on the data.

Results
Using Pearson's r correlation, the relations between all the variables were considerable.All the correlations, except one, were significant.They ranged from .20 (social reinforcement goals-learning goals) to .47 (performance goals-social reinforcement goals).It is notable that the degree of association between social and achievement goals was almost zero and statistically non-significant.
Table 2 presents the results obtained for the contrast of means for independent samples (t-test) in which the grouping variable was the degree of visual impairment (total or partial) and the grouped variables were the goal orientations.For social goals, partially sighted students obtained a higher mean score than blind students.A k-means cluster analysis was performed to establish students' profiles according to the scores they obtained for the goals.The method used was centroid linkage and the measure, Pearson's correlation, since the aim was to determine how subjects responded (Martínez-Arias, 1999).All the variables used in the clustering processes were relevant (Learning goals F 2; 168 = 136.27;p < .001;Social reinforcement goals F 2; 168 = 163.30;p < .001;Performance goals F 2; 168 = 29.72;p < .001;and Achievement goals F 2; 168 = 5.23; p = .006).
Table 3 shows the final cluster centroids for each variable.Students in the first cluster (n = 63) had high mastery goals, intermediate performance and achievement or reward goals and low social goals.Students in the second cluster (n = 52) presented high learning, social and performance goals and intermediate achievement goals.In the third cluster (n = 56), students' learning, social reinforcement, performance and achievement goals were below average.These three goal orientation profiles are depicted more clearly in Figure 1.  4 gives the results of the contrast of functions (Wilks' Lambda test) in both groups and the eigenvalues obtained.The model was constructed in three steps using the first two canonical discriminant functions in the analysis.For blind students, the first discriminant function explained 83.4% of the variance and the canonical correlation was .82,compared with 16.6% and .54for the second function.In other words, the profiles differed more with respect to the first function.For partially sighted students, the first discriminant function explained 80.5% of the variance and the canonical correlation was .81,compared with 19.5% and .56 for the second function, indicating that the profiles differed more with respect to the first function.Finally, a Discriminant Analysis was conducted using the goals as predictors and the cluster as predicted variables.This was done to depict the capacity of each of the goals to explain the variability due to differences between the three student profiles.The achievement goal was not used in the analysis.Table 5 shows the structure matrix of the Discriminant Analysis.The discriminant functions correctly classified 88.7% of blind students and 93.6% of partially sighted students.Note: + Greater absolute correlation between each variable and any discriminant function.

Discussion
This study had two objectives.The first was to determine whether goal orientations differed according to degree of visual impairment (total and partial).
The results obtained show a statistically significant inter-relationship between the four goal orientations, with the exception of social and achievement goals.These findings indicate that the goals were not orthogonal and that the same subject could pursue different goals at the same time, in line with the idea of multiple goals described by Wentzel (1993Wentzel ( , 1999Wentzel ( , 2000)).In addition, the social goals presented statistically significant differences, whereby partially sighted students obtained higher mean scores.It may be that their residual vision allows them to perceive the social and emotional reactions of others better, and they are therefore more motivated to obtain social approval.This is further supported by studies showing that people with total visual impairment present poor social functioning (Bieber-Sucht, 1991;Navas et al., 2017;Van Hasselt, 1983;Verdugo & Caballo, 1995;Warren, 1984), manifested in a chronological delay in the development of social skills, less social contact, the existence of problems regarding social adaptation and isolation, a lack of assertiveness and poorer interpersonal social skills.
The second objective was to establish motivational profiles as a function of goal orientations, as it has been obtained in other studies (Alonso-Tapia et al., 2010;Barca-Lozano et al., 2012;Barca et al., 2011;Cecchini et al, 2011;Pekrum et al., 2009).Three profiles were obtained consistent with those reported in other studies of students with visual impairment (Holgado et al., 2012;Jover, 2009).The first stands out for obtaining high scores in learning goals; the subjects of the second group are defined by obtaining high scores in learning goals, social reinforcement goals and performance goals; and the subjects of the third group are characterized by their low scores on learning goal, social reinforcement goals and achievement goals.
The discriminant analysis results indicate that for blind students, learning and social reinforcement goals were the most important variables in the second discriminant function, whereas academic performance goals were more important in the first function.However, for partially sighted students, learning and academic performance goals were the most important variables in the first discriminant function, while social goals were more important in the second function.In addition, the achievement or reward goals presented no discriminant power.
Taken together, this indicates that students were classified into one of the three profiles obtained as a function of different goals depending on their status (blind or partially sighted).
Exercising self-criticism, one of the main potential limitations of this study was that in reference to visual impairment (total or partial), no account was taken of when the impairment was acquired nor of its aetiology.It should be borne in mind that in terms of cognitive and social development, it is not the same to become blind after an accident as it is to be born blind; neither is the degree of acuity and visual field the same for all partially sighted people, since there is a vast difference between barely requiring any instrument to read and depending on assistive technology, as do some partially sighted people who are almost totally blind (Herrera & Ramírez, 1999;Navas et al., 2017;Rosa et al., 1993;Rosa et al, 1986;Van Hasselt, 1983;Verdugo & Caballo, 1995;Warren, 1984).
Nor did we consider students' sociocultural level, the quality of their affective relations at home, or their families' attitudes, which exert such a strong influence on students' perception of education.As Núñez (1999) and Pelechano et al. (1995) have noted, there is a close relationship between parents' attitudes and the levels of development attained by children with visual impairment but without other associated disorders.
Another important variable that was not considered was the age range of participants, from 8 to 27 years old.It appears that academic goals diversify with an increasing the level of education (Navas et al., 2002).Because we did not control for age or educational stage, it remains unclear in the present study whether the goal orientations studied were the result of visual impairment or, conversely, were determined by the different ages and educational stage of the participants.
Another critical issue is the use of a questionnaire of goals with a Likert scale, but "what do students say when asked to describe their goals in their own words?Remarkably, very few investigators have addressed this question" (Brophy, 2005, p. 170).It has been suggested that students rarely generate performance goals spontaneously and this is revealed when they are asked to describe their goals in their own words (Brophy, 2004;Navas et al., 2012).
In future research on the academic goals of people with visual impairment, these variables should be considered together with others such as learning contexts, attributions and self-concept.
In conclusion, the following findings should be highlighted: a) The only statistically significant differences obtained between blind and partially sighted students were for social reinforcement goals, with the latter group obtaining a higher mean score than the former.b) Three clusters or profiles were identified, which were used to classify the students according to the scores they obtained for the different goals.Of these, the ones that demonstrated the greatest discriminant power were learning goals, academic performance goals, and social reinforcement goals, although their discriminant capacity varied according to student status (blind or partially sighted).

Table 1
Summary of the results of the Exploratory Factor Analysis

Table 2
Group statistics based on visual impairment (total / partial) Note: VI = Visual impairment * p < .05

Table 3
The final cluster centroids

Table 4
Tests of equality of means for the groups and eigenvalues obtained

Table 5
Structure matrix