Page 26 - Kukanja Gabrijelčič, Mojca, and Maruška Seničar Željeznov, eds. 2018. Teaching Gifted and Talented Children in A New Educational Era. Koper: University of Primorska Press.
P. 26
odoros P. Kokkinos, Aikaterini D. Gari, and Lavrentios G. Dellassoudas
of Superior Students (SRBCSS) (Renzulli, Siegle, Reis, Gavin, & Reed, 2009;
Renzulli, Smith, White, Callahan, & Hartman, 2002).
During a pilot study, ten interviews with participants from schools of al-
most the whole range of Attica were conducted. Internal consistency of the
instrument’s items found to be at a very high level (Cronbach’s α = 0.90) in-
dicating high reliability.
Results
Descriptives
The highest mean was noted in the variable of Speech Accuracy (M = 4.5,
SD =.99), followed by Speech Adaptability to the Communicative Situation
(M = 4.4, SD = 1.26). Moreover, the means were high for the variable of Task
Commitment (M = 4.3, SD = 1.43) and the responsiveness to Problem Solving
Questions (M = 4.0, SD = 1.56). On the contrary, the Originality of Responses
(M = 3.5, SD = 1.86) and Non Conformity (M = 2.0, SD = 1.78) had the lowest
means.
Factor Analysis
In order to explore the correlations further, an exploratory factor analysis
with a Principal Components method was conducted. Prior to this, the fac-
torability of the 18 items was examined. Firstly, 15 of the 18 items correlated
at least 0.30 with at least one other item, suggesting reasonable factorability.
Secondly, the Kaiser-Meyer-Olkin measure of Sampling Adequacy was 0.758,
above the recommended value of 0.6, and Bartlett’s Test of Sphericity was
significant (χ2(153) = 640.646, p < 0.001). Initially, a five factor solution was
suggested by a Principal Components Analysis with direct oblimin rotation,
explaining 64.45 of total variance. However, three cross-loadings appeared
on the factor loading matrix. Thus, a three factor solution, which explained
51.33 of the variance, was preferred because of its previous theoretical sup-
port, the ‘levelling off’ of eigenvalues on the scree plot after three factors,
and the difficulty of interpreting the fourth and fifth factors. As can be seen
in Table 1, which presents the factor loading matrix for the final three factor
solution, the factor labels suited the extracted factors and correspond to the
three major item groups (linguistic, creativity, motivation) investigated in the
current study. Internal consistency for each of the scales was examined using
Cronbach’s alpha. The alphas were satisfactory, 0.818 for linguistic features (8
items), 0.72 for creativity (6 items), and 0.62 for motivation (3 items). No sub-
stantial increases in alpha for any of the scales could have been achieved by
eliminating more items.
24
of Superior Students (SRBCSS) (Renzulli, Siegle, Reis, Gavin, & Reed, 2009;
Renzulli, Smith, White, Callahan, & Hartman, 2002).
During a pilot study, ten interviews with participants from schools of al-
most the whole range of Attica were conducted. Internal consistency of the
instrument’s items found to be at a very high level (Cronbach’s α = 0.90) in-
dicating high reliability.
Results
Descriptives
The highest mean was noted in the variable of Speech Accuracy (M = 4.5,
SD =.99), followed by Speech Adaptability to the Communicative Situation
(M = 4.4, SD = 1.26). Moreover, the means were high for the variable of Task
Commitment (M = 4.3, SD = 1.43) and the responsiveness to Problem Solving
Questions (M = 4.0, SD = 1.56). On the contrary, the Originality of Responses
(M = 3.5, SD = 1.86) and Non Conformity (M = 2.0, SD = 1.78) had the lowest
means.
Factor Analysis
In order to explore the correlations further, an exploratory factor analysis
with a Principal Components method was conducted. Prior to this, the fac-
torability of the 18 items was examined. Firstly, 15 of the 18 items correlated
at least 0.30 with at least one other item, suggesting reasonable factorability.
Secondly, the Kaiser-Meyer-Olkin measure of Sampling Adequacy was 0.758,
above the recommended value of 0.6, and Bartlett’s Test of Sphericity was
significant (χ2(153) = 640.646, p < 0.001). Initially, a five factor solution was
suggested by a Principal Components Analysis with direct oblimin rotation,
explaining 64.45 of total variance. However, three cross-loadings appeared
on the factor loading matrix. Thus, a three factor solution, which explained
51.33 of the variance, was preferred because of its previous theoretical sup-
port, the ‘levelling off’ of eigenvalues on the scree plot after three factors,
and the difficulty of interpreting the fourth and fifth factors. As can be seen
in Table 1, which presents the factor loading matrix for the final three factor
solution, the factor labels suited the extracted factors and correspond to the
three major item groups (linguistic, creativity, motivation) investigated in the
current study. Internal consistency for each of the scales was examined using
Cronbach’s alpha. The alphas were satisfactory, 0.818 for linguistic features (8
items), 0.72 for creativity (6 items), and 0.62 for motivation (3 items). No sub-
stantial increases in alpha for any of the scales could have been achieved by
eliminating more items.
24
