The research behind Mindmore
The validity of Mindmore's tests has been evaluated in a study where each test was compared with the corresponding paper-based version. The study was conducted in collaboration with Stockholm University, Karolinska Institutet and the University of La Laguna. 81 healthy individuals aged 24–81 years participated. With an interval of 4 weeks on average between the test sessions and in a balanced order, the participants underwent two test sessions, one digital and one paper and pen-based, with the same tests. Correlations, t- and equivalence tests showed acceptable comparability between the formats.
Read Björngrim et al., (2019) published research article in Frontiers in Psychology .
In collaboration with the University of Skövde, a study was conducted to evaluate the test-retest reliability of Mindmore. The study included 40 healthy individuals who performed Mindmore's tests on two occasions, four weeks apart. Primary results from the study were (1) strong correlation between test sessions for 10 out of 14 tests and (2) median Pearson's r for all tests = 0.62, range [0.34, 0.83]. Further analysis of results is ongoing and the article presenting the study is in preparation.
Mindmore uses a regression-based approach to calculate test norms on an individual basis, and test-specific regression weights for age, level of education, and sex. Normative data was collected in collaboration with Stockholm University and Karolinska Institutet. Additional standard data is collected continuously.
In our normative study, a total of 720 healthy adults (17 to 93 years) completed the Mindmore screening battery, which consists of 22 individual (sub)tests across five cognitive domains: attention and processing speed, memory, language, visuospatial functions and executive functions. Regression-based normative data were established for 42 test result measures, investigating linear, non-linear and interaction effects between age, education and sex.
The test results were most affected by age and to a lesser extent by education and sex. All but one test displayed either linear or accelerated age-related decline, or a U-shaped association with age. All but two tests showed beneficial effects of education, either linear or subsiding after 12 years of educational attainment. Sex affected tests in the memory and executive domains. In three tests, an interaction between age and education revealed an increased benefit of education later in life.
This study provided normative models for 22 traditional cognitive (sub)tests adapted for self-administration through a digital platform. The models enables more accurate interpretation of test results, hopefully leading to improved clinical decision making and better care for patients with cognitive impairment.
Figure 1. Regression based model for one of the (sub)tests (SDMT)