Table step one shows the fundamental detailed analytics into the examined circumstances and also the IDI (Goods Difficulty Index) value
It was not noticed that distribution away from examined situations differed rather on the regular delivery. Concurrently, IDI index didn’t confirm the presence of floors or roof outcomes on studies (simply in the example of items nine “I continue trying to find anything” performed a small ceiling effect let you know for all about three products). Frequency analysis to possess private test activities displayed no troubles having investigation granulation. In that respect, most of the points were included in subsequent analyses. It because of effect place bias, including social desirability and you can acquiescence.
Factorial Legitimacy
By applying the Velicer MAP method, it was found that the optimal number of components to be extracted is one, which confirms the original factor structure of the questionnaire. The effect was observed both when the results of the interviewees were analyzed together and when they were divided into validation groups. Subsequently, the result obtained by MAP was verified by means of principal component analysis with cougar life oblimin rotation. As the criterion for the number of factors to be isolated, the eigenvalue was set to be equal to at least 2. Assumptions of factor analysis were met [KMO = 0.908; chi ( 91 ) 2 = 1; p < 0.001]. The created factor explains a total of % variance of the questionnaire. The factor structure was checked for the three validation groups, there were no significant differences in the size of explained variance: early adulthood (%), young adolescence (%), problem group (%). The obtained factorial results correspond to the expected one factor dimensions of the original version of the Questionnaire.
Parallel data indicated that a-one-factor solution is the most likely for everyone about three brand new examples. Given that indicated in Desk dos, the one-basis model demonstrated an excellent item loadings and you may equivalent abilities have been noticed regarding the product loadings over the around three examples. Every activities loaded saliently with this grounds, that have item packing varied anywhere between 0.505 and 0.719 during the early adulthood take to. Throughout the teenagers it actually was anywhere between 0.464 and you will 0.770 plus the trouble category it had been ranging from 0.400 and 0.668. A reduced loadings remained significant, verifying the fresh new coherent design of one’s Polish brand of the latest means.
The next phase provided contrasting the outcome of your own design fitting tips off their available studies. There have been no significant differences when considering the fresh design fitting score received regarding the confirmatory grounds studies (CFA). The brand new Polish sort of RS-14 displayed a coherent one-dimensional factor build with reples.
As well, it needs to be indexed one to on the about three sample recognition groups no biggest differences had been observed. Detailed results are provided during the Dining table step three.
Cross-Classification Legitimacy Precision out-of Scales
Cronbach’s leader coefficient was utilized to look at the internal consistency out of the brand new obtained survey evidence. The outcomes is actually showed below.
Brand new accuracy of one’s created foundation are 0.853 (total try) regarding validation trial, hence verifies its higher texture. Immediately following section with the recognition organizations, zero tall variations in level consistency was in fact observed. Cronbach’s alphas varied ranging from 0.824 and you may 0.871 along the around three products:
The best precision of one’s received foundation is actually by young people (? = 0.871) and you may respectively lower of the condition classification: (? = 0.824).
Therefore, the internal consistency reliability of the RS-14 is not only acceptable across all sample populations, but also rather robust. Additionally, the analysis of the discriminative power of individual items shows that all test items are positively correlated with the scale. Because of the slight variation in Cronbach’s Alpha value for items where the correlation with the scale was <0.300, it was decided to include all test items in the factor. The exact results are shown in Table 4 below.