Page 147 - FULL REPORT 30012024

Table 5.6 Summary of Behavioural Intention Responses
No Questions Respondents count Average

in each Score
1 2 3 4 5
1 I intend to use to check my 0 0 6 8 16 4.13
stroke risk and to refer
information on stroke.

2 I would recommend this 0 0 4 10 16 4.33
application to others interested
in assessing their stroke risk
and to understand more about
stroke.

The first question garnered an average rating of 4.13, with no participants

giving it the lowest ratings of 1 or 2. The majority of 24 out of 30 respondents
rated it as 4 or 5, suggesting a favourable tendency towards using the

programme for health monitoring. The second question yielded a higher

average score of 4.33, indicating a significant inclination among users to
endorse the programme to others. The application's potential for broader

appeal is shown by the fact that 26 out of 30 respondents did not provide poor
ratings (1 and 2) or high ratings (4 and 5).

5.3 Findings

The findings from the prediction model evaluation shows that overall, the

stroke risk prediction model demonstrates a strong performance, particularly

excelling in its ability to identify actual cases of stroke risk, as evidenced by
its high recall rate of 90%. This aspect is crucial especially for medical

contexts were failing to detect a true positive can have serious implications.
The lowest score, which is the precision of the model, at 75%, also still

signifies a relatively strong performance level which indicates that three-

quarters of the positive stroke predictions made by the model are accurate.
Such a rate is noteworthy in the healthcare sector, as it demonstrates the

model's substantial capability in correctly identifying stroke risks. The

130