4.1 General analysis of the results
4.1.1 Gender analysis:
This work allows to analyse the perception of people with disabilities regarding the access and use of technology. A survey was used as a working method, which was completed by 102 people, including 50 women and 52 men.
4.1.2 Age analysis:
In terms of age, it has been observed that the age range has been variable in both sexes:
Graph 2 shows that the most predominant age group in the case of women is 26-35 and 36-45 years, followed by 18-25 years and then 46-55 years. It is also observed that few people over 56 years of age have answered the survey.
As for men, the group that predominates the most is that of 36-45 years, followed by 18-25 years and 26-35 years with the same percentage. It is observed, like women, that there are few men who are over 46 years old.
4.1.3 Type of disability analysis:
Among the people surveyed, the following differences have been highlighted in terms of the disabilities they suffer:
Graph 3 shows that intellectual disability is the most predominant with a percentage of 32% in the case of women and 46% in the case of men, followed by visual and physical disability at 18% each for women. and 15% and 12% respectively for men. This graph also highlights that there are people who suffer from several disabilities at the same time.
4.1.4 The daily problems they face
In the following graph, the results of the daily problems that people with disabilities face in their day to day are presented, the answers follow the concept of a Likert scale that goes from 1 to 5 where 1 means very little and 5 very frequent. This scale is used too for the graph 6,7,8.
Graph 4 shows that women are the ones who have scored the most with 5 to the problems they face in their day to day and men who have scored the most with 1 to these difficulties encountered. Among the problems with the highest score (5) in the case of women are access to adequate employment (50%) and carrying out administrative procedures (48%), while for men the rate of access problems to adequate employment represents 37% and is the highest rate of the problems indicated, followed by that of administrative procedures, which represents 29%.
4.1.5 The availability of ICT tools
This graph shows the technological tools that our sample has by gender:
Graph 5 shows that the Internet and the mobile phone are the two most widely available tools for both men and women, unlike the other tools where the difference in percentages of availability is more marked with predominant use by men.
On the other hand, it is highlighted that the rate of women who do not have technological tools, but would like to have them, is higher than that of men. Among them, the printer, the antivirus, the hard drive, the video conference program, the smartwatch stands out for women, while for men they are the printers, the hardware drive and the smartwatch. Also, it can be detected that specific technological tools for people with disabilities is a more important for women than for men, since this rate represents double (64% compared to 31%). With this global result, we observe that the rate of availability of all technological tools is higher in the case of men. This can be justified by the higher employment rate of men compared to women and, in fact, by having more financial resources than women.
4.1.6 Ability to use technology
Question 6 of the survey asks if people with disabilities know how to use certain technological tasks to solve certain procedures. The question uses a Likert scale from 1 to 5, where 1 means very poorly and 5 means very well.
Graph 6 shows that women consider themselves less capable of handling administrative procedures, searching for a GPS location, requesting a medical appointment, using banking services and buying online, as well as requesting a home delivery service. It is the reason why they majority gave the punctuation of 1 to these tasks, while they considered impartially capable of sending email, searching for information on the Internet and using social networks, where the highest score is 3. On the contrary, men have highlighted their ability to use ICTs, such as sending email, searching for information on the Internet, locating a site on GPS, doing medical management, and using social networks, where the highest response percentage is 5, while who have given the highest score to 1 in administrative management, making use of banking services, making an online purchase or sale, and requesting home service.
Therefore, it is observed that the percentages of scores with 5 of the men in terms of abilities to use ICTs are almost all higher than those of the women, except for "requesting home services" with a difference of 1%, but it is also noted that they did not exceed 50%, which is why a dispersion is observed in the graph. While the percentages of scores with 1 for women are all higher than those for men, but all, except one, are around 60 %, which indicates a great concentration in the graph. This result is closely related to the previous one, since it leads us to deduce that the fact of not discarding technological tools in the same way as men means that the ability to use them is not the same.
4.1.7. Difficulty in using technology
Question 7 of the survey aims to determine the difficulties that people with disabilities have in the use of information and communication technologies. In this case, a Likert scale from 1 to 5 has also been proposed, where 1 means totally disagree and 5 means totally agree.
Graph 7 shows that both men and women have given the maximum score (5) to all the difficulties mentioned. However, the percentages are different, since we can see that 54% of women have indicated that they find it difficult compared to 37%, that 72% of women have indicated that technology costs a lot of money compared to 60% of men, that 54% of women indicate that ICTs are not adapted to their disability compared to 37% of men, that 50% of women indicate that ICTs generate distrust compared to 29% of men and 72% of women indicate that they need specific training compared to 50% of men.
These results are in line with what was previously indicated by the United Nations [13] and Epdata [22], where it is indicated that men with disabilities have greater access to education, the labour market, and higher salaries. This reflects the difficulty that women have in acquiring the tools due to their high cost, the need for training to become familiar with the technology and the difficulty of using them.
Also, they are a consequence of what was previously observed in graph 5, where we saw that women indicate that they need more specific technologies for their disabilities compared to men, and that is the reason why they indicate that the difficulty they have is due to the non-adaptability of ICTs to their needs.
4.1.8. Impact of technology on the lives of people with disabilities
Finally, the sample was asked how technology has affected their lives. Like the previous questions, a Likert scale ranging from 1 to 5 has been proposed, where 1 is “totally disagree” and 5 is “totally agree”.
In graph 8, the score of the contribution of ICTs to men is higher than that of women in all points, which is reflected in the graph in a greater concentration on the right side for men. This result indicates that having been able to have technological tools and knowing how to use them has been enjoyed more of his use and consequently several things have been improved in the life of men compared to women. As a result of the analysis of these graphs [4-8], it is verified that the gender variable affects the digital gender gap in the case of people with disabilities, which leads us to accept H1 and reject H0, leaving pending to determine if the disability affects the digital divide in the case of people with disabilities (H2), an issue that is analysed later.
4.2 Analysis of the mean, variance, standard deviation, and Cronbach's alpha of each of the categories
To evaluate the reliability of the results, a study of different statistical elements such as means, variances, deviations and Cronbach's Alpha has been carried out:
Table 1. Análisis estadístico de las variables
|
Mean
|
Variance
|
Deviation
|
Cronbach's Alpha
|
Daily problems
|
3.28
|
1.56
|
1.25
|
0.92
|
ICT capabilities
|
2.79
|
1.38
|
1.18
|
0.94
|
Difficulty in using ICTs
|
3.78
|
1.40
|
1.18
|
0.87
|
Impact of ICTs
|
3.51
|
0.98
|
0.99
|
0.87
|
Source: Own elaboration
The results of table 1 show that the means of the four results evaluated by a Likert scale reflect a mean around 3, with a deviation of plus or minus 1 to 1.25. To further emphasize this result, it is highlighted that the variance of the survey indicates that the daily problems present a greater variance, therefore, the dispersion of the distribution of the opinion of the surveyed population is greater than in the rest. While the impact of ICTs represents the least variation and consequently the least deviation of opinions.
Regarding the Cronbach's Alpha indicator, it indicates that our results follow a reliable logic, since the reliability indices are high (between 87% and 94%).
It should be remembered that access to technology is not considered in this analysis, because it is not a numerical variable but rather a categorical one.
4.3 Average by gender and disability
4.3.1 Average by gender
To evaluate the differences by gender in terms of the five elements, Table 2 is presented:
Table 2. Mean by gender
|
Daily problems
|
Access to ICTs
|
ICT capabilities
|
Difficulty in using ICTs
|
Impact of ICTs
|
Yes
|
No, I don't need it
|
No, but I would like to
|
Men
|
2.96
|
0.61
|
0.14
|
0.25
|
3.12
|
3.48
|
3.56
|
Women
|
3.61
|
0.43
|
0.48
|
0.09
|
2.44
|
4.10
|
3.44
|
Confidence interval
|
[-1.12, -0.18]
|
[0.08, 0.28]
|
/
|
/
|
[0.23, 1.12]
|
[1.07, 0.17]
|
[-0.26, 0.52]
|
Source: Own elaboration
Table 2 shows very visible differences between the two genders, since it is observed that the average score of men in terms of daily problems is lower than that of women: 2.96 compared to 3.61 for men, which indicates that men have fewer daily problems than women in a confidence interval that varies between [-1.122, -0.18].
The same is observed in terms of access to ICTs, in which the availability rate of ICTs on average is 0.61 in the case of men and 0.43 in the case of women, which shows that men have greater access to ICTs within a confidence interval of [ 0.08, 0.28].
Regarding the use of ICTs, the mean score of men is higher than that of women (3.12 compared to 2.44), which indicates that men have a greater use of ICTs in a confidence interval of [0.23, 1.12].
Regarding the difficulty of using ICTs, it is observed that women are the ones who have scored this difficulty the most: 4.10 compared to men 3.48, which indicates that they are the ones with the greatest difficulties in using ICTs at a confidence interval of [-1.07, -0.17].
The impact of ICTs on the lives of people with disabilities shows a difference in favour of men, although in this case the difference is not significant (3.56 versus 3.44), which shows that men have greater impact than women within a confidence interval of [-0.26, 0.52].
4.3.2 Average per disability
In this section we want to measure the impact that the type of disability has on the digital divide to validate hypothesis H2, for which the post-hoc analysis of Tukey and Benferroni is used, which consists of finding honest significant differences to obtain the confidence intervals in the case of the means and the corrected P values that allow accepting or rejecting the test. For this reason, the results are proposed in table 3:
Table 3. Multiple comparison of the means of the types of disabilities -Tukey analysis and Bonferroni -
Groups
|
|
Multiple Comparison of Means with Tukey Model
|
|
Multiple Comparison of Means with Benferroni Model
|
group1
|
group2
|
meandiff
|
p-adj
|
lower
|
upper
|
reject
|
Stat
|
pval
|
reject
|
Auditory
|
Physical
|
-0.69
|
0.60
|
-1.91
|
0.53
|
False
|
1.84
|
0.08
|
False
|
Auditory
|
Intellectual
|
-0.16
|
0.90
|
-1.21
|
0.90
|
False
|
0.43
|
0.67
|
False
|
Auditory
|
Multiple
|
-0.46
|
0.90
|
-1.79
|
0.88
|
False
|
1.16
|
0.26
|
False
|
Auditory
|
Organic
|
-0.01
|
0.90
|
-1.65
|
1.62
|
False
|
0.04
|
0.97
|
False
|
Auditory
|
Mental health
|
-0.79
|
0.75
|
-2.42
|
0.85
|
False
|
1.60
|
0.14
|
False
|
Auditory
|
Visual
|
-0.19
|
0.90
|
-1.38
|
0.99
|
False
|
0.61
|
0.55
|
False
|
Physical
|
Intellectual
|
0.53
|
0.56
|
-0.37
|
1.43
|
False
|
-1.64
|
0.11
|
False
|
Physical
|
Multiple
|
0.23
|
0.90
|
-0.98
|
1.45
|
False
|
-0.55
|
0.58
|
False
|
Physical
|
Organic
|
0.68
|
0.81
|
-0.86
|
2.21
|
False
|
-1.39
|
0.18
|
False
|
Physical
|
Mental health
|
-0.10
|
0.90
|
-1.63
|
1.44
|
False
|
0.17
|
0.87
|
False
|
Physical
|
Visual
|
0.50
|
0.76
|
-0.55
|
1.56
|
False
|
-1.52
|
0.14
|
False
|
Intellectual
|
Multiple
|
-0.30
|
0.90
|
-1.35
|
0.75
|
False
|
0.79
|
0.43
|
False
|
Intellectual
|
Organic
|
0.14
|
0.90
|
-1.27
|
1.56
|
False
|
-0.29
|
0.77
|
False
|
Intellectual
|
Mental health
|
-0.63
|
0.81
|
-2.04
|
0.78
|
False
|
1.21
|
0.23
|
False
|
Intellectual
|
Visual
|
-0.03
|
0.90
|
-0.89
|
0.83
|
False
|
0.11
|
0.91
|
False
|
Multiple
|
Organic
|
0.44
|
0.90
|
-1.19
|
2.07
|
False
|
-0.90
|
0.39
|
False
|
Multiple
|
Mental health
|
-0.33
|
0.90
|
-1.96
|
1.30
|
False
|
0.55
|
0.60
|
False
|
Multiple
|
Visual
|
0.27
|
0.90
|
-0.92
|
1.46
|
False
|
-0.75
|
0.46
|
False
|
Organic
|
Mental health
|
-0.77
|
0.87
|
-2.66
|
1.11
|
False
|
1.29
|
0.23
|
False
|
Organic
|
Visual
|
-0.17
|
0.90
|
-1.69
|
1.34
|
False
|
0.44
|
0.66
|
False
|
Mental health
|
Visual
|
0.60
|
0.90
|
-0.92
|
2.11
|
False
|
-1.28
|
0.22
|
False
|
Source: Own elaboration
Table 3 consists of a multiple hypothesis test to compare the means of the pairwise disabilities. Therefore, it is structured in two columns (group 1 and group 2), to contemplate all the possible combinations of disabilities. For example, auditory disability is compared with all other disabilities, the same happens with physical, intellectual, visual, organic, and mental, which represent all the existing disabilities in this study. Adding to these, the multiple disability, which represents the combination of two or more capacities.
In the first Tukey analysis, the difference of the means (Meandiff), the p value or the adjusted value is presented, which must be less than 0.05 to valid the comparative, and finally the lower and upper values that represent the confidence intervals of the value.
According to Tukey's analysis, to accept the hypothesis, the values of the Lower and Upper intervals have to appear with the same sign (either both positive or both negative). In this case, the hypothesis is rejected since these two reference elements have different signs, which means that zero is not included, which is showed also in the column seven by the sign of “False” in all the lines.
To be sure of this result and to validate it, the analysis has been repeated using the Benferroni model, where it has been seen that the hypothesis that the type of disability affects the digital gender gap is also rejected. This is seen by the value of the p that are all greater than 0.05, which is showed also in the column ten by the sign of “False” in all the lines. In this table, the Stat value is also observed, which is a method used for testing and adjustment of p values and consists of measuring the difference in means between the two groups.
4.4. Correlation analysis of quantitative variables and multivariate regression
Finally, and to see the correlation between our five variables, the general correlation matrix is presented below in Figure 9:
This matrix shows that there is a high correlation (68%) between the “ICT capabilities” and the “impact of ICTs”, which justifies the differences previously studied in graphs 6 and 8.
A positive correlation of 65% is also observed between the “difficulty in using ICTs” with “daily problems” of people with disabilities. Almost equal to the existing correlation between the “Access to ICTs-No, but I would like to”, with the “difficulty in using ICTs” (61%), which is logical, given that if the tool is not available, it is not possible to avoid the difficulty of using it.
Also, a 55% correlation is observed between “Access to ICTs-Yes” with “ICT capabilities”. The correlation is not very high since if there has been no education or training in the use of the tool, even if it is available, it cannot be used properly.
On the other hand, there is a correlation (49%) between “access to ICTs-Yes” and the “impact of ICTs”. This correlation is positive but less impressive than that of use, since having technology without knowing how to use it, does not affect the lives of people with disabilities, in the same way, as having it and knowing how to use it. To summarize, it is presented in the following table:
Table 4. Comparison of the correlation coefficients and the p-value
Variables
|
Correlation coefficients (Pearson)
|
P-value
|
Daily problems - ICT capabilities
|
-0.31
|
1.4123 e-03
|
Daily problems - Difficulty in using ICTs
|
0.65
|
1.6306 e-13
|
Daily problems - Impact of ICTs
|
-0.22
|
2.3823 e-02
|
Daily problems – Access to ICTs-Yes
|
-0.47
|
7.9314 e-07
|
Daily problems - Access to ICTs- No, I don´t need it
|
-0.17
|
9.1647 e-02
|
Daily problems - Access to ICTs- No, but I would like to
|
0.52
|
1.6048 e-08
|
ICT capabilities - Difficulty of in using ICTs
|
-0.38
|
8.2283 e-05
|
ICT capabilities – Impact of ICTs
|
0.68
|
3.9957 e-15
|
ICT capabilities – Access to ICTs-Yes
|
0.55
|
2.1226 e-09
|
ICT capabilities – Access to ICTs- No, I don´t need it
|
0.08
|
3.9577 e-01
|
ICT capabilities – Access to ICTs- No, but I would like to
|
-0.54
|
3.4332 e-09
|
Difficulty in using ICTs – Impact of ICTs
|
-0.11
|
2.7225 e-01
|
Difficulty in using ICTs – Access to ICTs-Yes
|
-0.48
|
3.9996 e-07
|
Difficulty in using ICTs – Access to ICTs- No, I don´t need it
|
0.29
|
3.5149 e-03
|
Difficulty in using ICTs – Access to ICTs- No, but I would like to
|
0.61
|
8.8904 e-12
|
Impact of ICTs – Access to ICTs-Yes
|
0.49
|
2.1131 e-07
|
Impact of ICTs – Access to ICTs- No, I don´t need it
|
-0.11
|
2.8539 e-01
|
Impact of ICTs – Access to ICTs- No, but I would like to
|
-0.36
|
1.8299 e-04
|
Access to ICTs-Yes – Access to ICTs- No, I don´t need it
|
-0.19
|
5.9635 e-02
|
Access to ICTs-Yes – Access to ICTs- No, but I would like to
|
-0.77
|
7.8842 e-21
|
Access to ICTs- No, I don´t need it – Access to ICTs- No, but I would like to
|
-0.49
|
1.7975 e-07
|
Source: Own elaboration
This table resume the correlation coefficients of Pearson observed previously in the graph 9 and shows that the “ICTs capabilities” is the higher correlated variable with the “impact of ICTs” (68%), with a p-value of 3.9957e-15 which is lower than 0.05.