166 Chapter V Results and Discussions V.1 Introduction Within this chapter, computations and discussions shall be undertaken predicated upon data acquired from the field during the research phase. The analytical framework employed herein encompasses both descriptive analysis and empirically substantiated analysis via the application of Structural Equation Modeling (SEM) utilizing the LISREL software. Notably, path analysis is deployed for the quantification of constructs encompassing Digital Culture, General Culture, Digital Literacy, Attitudes to change, and Perceived Performance. It is imperative to emphasize that the study's sample population comprises a total of 383 individuals. V.2 Results V.2.1 Descriptive This section offers a descriptive overview of the respondents involved in the study. An examination of the profiles of the 383 participants reveals their distribution across various divisions and levels. Over half of the respondents are from the Consumer Division (CFU Consumer), with the remainder spread across other divisions, mirroring the overall distribution of Telkom employees (refer to Appendix 1 for details). Regarding job profiles, about 70% of the respondents are staff level employees (Band VI), with a decreasing number at higher levels. This distribution reflects the general position level composition of Telkom's workforce. Additionally, a detailed analysis of each variable has been conducted, revealing differences in the average scores for conscious and unconscious aspects (detailed findings can be found in Appendix 1). In this section, the respondent profiles based on age and the average values per variable for each group of respondents based on their birth generation will be specifically presented. This aims to further explore the existence of a generation gap and to set the stage for the discussion of intervention steps that will be included in this dissertation. 167 Table V.1 Respondents profile based on generation No Generation f % 1 1946-1964 (Baby boomer) 0 0 2 1965-1980 (Gen X) 46 12.0 3 1981-1999 (Gen Y) 335 87.5 4 2000-now (Gen Z) 2 0.5 Total 383 100.0 Based on the table above, it can be observed that the majority of respondents, specifically 329 individuals or 85.9%, fall into the age category of 1981- 1999 (Gen Y). Table V.2 Descriptive Analysis of Variables Based on Generation (Conscious) VARIABLE The average values for each generation 1946- 1964 1965- 1980 1981- 1999 >2000 All GENERAL CULTURE (GC) 0 4.34 4.27 4.12 4.27 Solid 0 4.62 4.41 4.00 4.43 Smart 0 4.09 4.02 4.00 4.02 Competent 0 3.93 3.91 3.34 3.91 Loyal 0 4.37 4.09 3.67 4.13 DIGITAL CULTURE (DC) 0 3.78 3.73 3.36 3.73 Learning Agility 0 4.54 4.57 4.00 4.57 Visionary 0 4.28 4.21 3.34 4.21 Willingness to Take Risks 0 3.61 3.63 3.84 3.63 Passion For Trial and Error 0 2.74 2.68 2.67 2.69 Empowerment 0 3.96 3.75 3.00 3.77 Collaboration 0 3.57 3.52 3.34 3.52 DIGITAL LITERACY (DL) 0 4.34 4.57 4.00 4.54 Operational 0 4.43 4.52 4.00 4.51 Information Navigation 0 4.43 4.68 4.00 4.64 Social 0 4.36 4.41 4.00 4.40 Creative 0 3.91 4.44 4.00 4.37 Mobile 0 4.57 4.79 4.00 4.76 ATTITUDES TO CHANGE (ATT) 0 4.44 4.46 3.95 4.45 Cognitive 0 4.64 4.58 4.00 4.58 Affective 0 4.17 4.28 3.84 4.27 Conative 0 4.50 4.52 4.00 4.51 168 Table V.2 Descriptive Analysis of Variables Based on Generation (Conscious) (cont.) VARIABLE The average values for each generation 1946- 1964 1965- 1980 1981- 1999 >2000 All PERCEIVED PERFORMANCE (PP) 0 4.30 4.27 3.67 4.27 Perceived Individual Performance 0 4.22 4.28 3.00 4.27 Perceived Team Performance 0 4.32 4.23 4.00 4.24 Perceived Organization Performance 0 4.36 4.30 4.00 4.31 Table V.3 Descriptive Analysis of Variables Based on Generation (Unconscious) VARIABLE The average values for each generation 1946- 1964 1965- 1980 1981- 1999 >2000 All GENERAL CULTURE (GC) 0 1.83 1.99 1.00 1.97 Solid 0 2.02 2.09 1.00 2.08 Smart 0 1.78 2.39 1.50 2.32 Competent 0 2.09 1.67 1.00 1.71 Loyal 0 1.46 1.75 1.00 1.71 DIGITAL CULTURE (DC) 0 1.83 1.81 1.00 1.81 Learning Agility 0 1.41 1.74 1.00 1.70 Visionary 0 2.17 1.99 1.00 2.01 Willingness to Take Risks 0 1.72 1.61 1.00 1.62 Passion For Trial and Error 0 2.04 2.02 1.50 2.02 Empowerment 0 2.15 1.95 1.00 1.97 Collaboration 0 2.72 2.77 1.00 2.75 DIGITAL LITERACY (DL) 0 2.15 2.03 1.00 2.04 Operational 0 1.63 1.66 1.00 1.65 Information Navigation 0 1.96 1.90 2.50 1.91 Social 0 2.74 2.83 1.00 2.81 Creative 0 2.11 1.96 1.00 1.97 Mobile 0 2.80 2.61 1.00 2.62 ATTITUDES TO CHANGE (ATT) 0 1.07 1.05 1.00 1.05 Cognitive 0 1.09 1.11 1.00 1.10 Affective 0 1.07 1.04 1.00 1.04 Conative 0 1.11 1.12 1.00 1.11 169 Table V.3 Descriptive Analysis of Variables Based on Generation (Unconscious) (cont.) VARIABLE The average values for each generation 1946- 1964 1965- 1980 1981- 1999 >2000 All PERCEIVED PERFORMANCE (PP) 0 1.26 1.22 1.00 1.22 Perceived Individual Performance 0 1.13 1.07 1.00 1.08 Perceived Team Performance 0 1.48 1.67 1.00 1.64 Perceived Organization Performance 0 1.41 1.16 1.00 1.19 Based on the cross- tabulation above, it can be observed that there is no difference in the average values for each variable across age generations. These findings lead to a more in-depth analysis through difference test for each of variables based on age. The difference test is conducted using ANOVA test among more than two populations on conscious data, which in this study is based on four age categories for each variable. Hypothesis Test : Ho: µ1 = µ2...= μ10 (There are no differences in each variable based on age) Ha: µ1 ≠ µ2…≠ μ10 (There are differences in each variable based on age) Test Criteria: Based on probability value:  If the value is Sig. ≥ 0.05, Ho accepted .  If the value is Sig. < 0.05, Ho rejected . 170 Table V.4 Results of ANOVA test Variabel F (Anova) p-value (Sig.) Conclusion GC 1.483 0.228 Supported DC 2.205 0.112 Supported ATT 0.123 0.884 Supported DL 2.001 0.137 Supported PP 0.553 0.575 Supported Based on the Anova test results in the table above, it can be seen that the p-value (Sig.) for each variable is greater than 0.05, indicating that Ho is accepted. Therefore, it can be concluded that there is no difference in each variable based on age. During the digital transformation, measuring conscious and unconscious behaviors yielded an interesting finding: there were no significant differences between generations in various aspects of corporate culture and adaptation to digital transformation. This result indicates that age or generation factors do not play a dominant role in measuring behaviors related to legacy culture, digital culture, digital literacy, attitudes to change , and performance assessment in the context of digital transformation. It suggests that the challenges in facing digital transformation are might be influenced by other factors such as experience, training, or individual changes rather than age or generation factors within the organization. These findings provide valuable insights into the importance of promoting corporate culture, digital literacy, and attitudes across generations in digital transformation efforts. V.2.2 Test Classical Assumptions This study has carried out classical assumption tests with results that meet the requirements according to the explanation in chapter III.3.5.4. This classic assumption test includes the data normality test, homoscedasticity test, and autocorrelation test, all of which show that the variables are normally distributed, that there is no heteroscedasticity, and there is no multicollinearity in the data. Thus, the results of this classical assumption test provide a strong basis for interpreting research results in further testing (See Appendix 2). 171 V.2.3 Independent T Test between Conscious and Unconscious Behavior Independent sample T test or two-group difference test was used to test two averages of two independent data groups (Prayitno, 2014). According to Ghozali (2018), the purpose of the independent sample T test is to be able to compare two groups that are not interconnected on conscious data. Hypothesis: Ho: µ1 = µ2 (There are no differences score between conscious and unconscious) Ha: µ1 ≠ µ2 (There are differences score between conscious and unconscious) α = 5% Test Criteria based on probability value: If the value is Sig. ≥ 0.05, Ho accepted. If the value is Sig. < 0.05, Ho was rejected. Table V.5 Independent T Test Results between Conscious and Unconscious Variabel T hitung p-value (Sig.) Conclusion GC 64.645 0.000 Supported DC 59.267 0.000 Supported ATT 142.976 0.000 Supported DL 68.418 0.000 Supported PP 96.006 0.000 Supported The table shows that the p-values for all variables are below 0.05, leading to the rejection of Ho. This means there are differences in general culture, digital culture, attitudes to change, digital literacy, and perceived performance between Conscious and Unconscious behavior. This difference arises because when respondents are tested using the Thematic Apperception Test (TAT) and respond by narrating their daily activities as per the provided images, they unconscious ly tend to use fewer words related to the twenty- one dimensions of measurement. This leads to machine learning predicting their behavior with an average score lower than that predicted based on their 172 questionnaire responses. According to theories proposed by Pennebaker (2013) and Ginting (2018), language serves not only as a communication tool but also reflects an individual's cognitive and emotional processes. The more words used in TAT stories can provide deeper insights and reflect a person's unconscious thoughts and feelings. Here, the more a respondent's story includes words related to the twenty- one research model dimensions, the higher the assessment score, and vice versa. TAT produces more spontaneous and emotional language pattern responses, which may more accurately describe unconscious attitudes and emotions related to digital transformation. These responses allow for the capture of deeper data and psychological nuances that respondents might not be fully aware of. In describing stories from the images given in TAT, respondents unconscious ly bring up words that reflect themselves. Additionally, machine learning text analysis can identify patterns and themes that might not be immediately apparent, offering broader insights into unconscious attitudes and emotions. On the other hand, questionnaires generate more conscious responses, possibly influenced by the respondent's desire to provide answers that are considered "correct" or socially acceptable. Employees provide direct and structured data through consciously chosen answers to multiple-choice questions, making questionnaire responses limited by respondent awareness and self-consciousness. Responding in this manner tends to result in relatively higher measurement scores. In conclusion, average behavior assessment scores reflected in the words of TAT stories analyzed with machine learning may provide a more accurate depiction of employees' unconscious mindsets and behaviors compared to the more conscious answers of a questionnaire.