zoom controls
datadata summary

Data coding in SPSS

Research groups and subgroups are designated according to Dylsexia Index boundary values - rationale detailed here and on the project StudyBlog. For quick reference, the table below indicates these groups and subgroups and their respective coding label in SPSS:

Research Group SPSS category code Research SubGroup SPSS category code Criteria
ND 0 ND-400 10 students in research group ND who present a Dyslexia Index Dx < 400
NDx400 90 students in research group ND who present a Dyslexia Index 400 < Dx <592.5
DNI 20 students in research group ND who present a Dyslexia Index Dx > 592.5 - this is the group of greatest interest
DI 1 DI-600 21 students in research group DI who present a Dyslexia Index Dx > 592.5 - this is essentially my 'control' group
DIx600 91 students in research group DI who present a Dyslexia Index Dx <592.5


Data summary

The first table (below) is the dataset summary built from 166(good)/183 datasets received. the 17 'spoiled' datasets were discarded because the questionnaires submitted by those respondents were substantially incomplete.

The summary table shows data for Dyslexia Index (Dx) and corresponding Academic Behavioural Confidence (ABC) for each of the 5 research subgroups (see explanatory notes below). The most significant feature that emerges from this summary table is the comparison of Academic Behavioural Confidence levels for the two most interesting research subgroups: that is group 'DNI' - student with no previously identified indication of dyslexia, and group 'DI-600' - students with an identified dyslexia and a Dyslexia Index of Dx > 592.5.
The Effect Size difference reported at the foot of the summary table is Hedges' 'g' for Academic Behavioural Confidence, as is the independent samples means t-test result.
These results together are indicating a moderate effect size between ABC for these key, research groups together with a statistically significant difference at the 5% level between the sample mean ABC values.

More comprehensive summary tables will be presented here in due course. A deeper critique is currently available on the StudyBlog, particularly discussing internal consistency reliability (Cronbach's α) and the Principal Compenent Analysis conducted on the metrics ABC and Dx

results summary
Explanatory notes:

  • eQNR responses self-sorted into the two, primary Research Groups:
    • DI - students who identified themselves as presenting Dyslexia. This is RG:DI and 68 eQNR responses were received.
    • ND - students indicating that they had no learning challenges or had learning challenges other than dyslexia. This is RG:ND and 98 eQNR responses were received.
  • Research Group: DNI is a subgroup of RG:ND and these students were identified through the Dyslexia Index (Dx) value that their replies generated. The boundary point for skimming these responses was Dx = 600.
  • A second subgroup of RG:ND has been established and this is students who presented a Dyslexia Index of Dx < 400. This subgroup is labelled RG:ND-400. Comparisons will be made against this subgroup as students presenting this low Dx value can be safely considered as presenting very low levels of dyslexic attributes - that is, they do not present a dyslexic learning difference.
  • Datasets from RG:DNI will be compared with datasets from RG:DI which present a Dyslexia Index of Dx > 600 (and this subgroup of RG:DI is designated RG:DI-600), and datasets in RG:ND-400.

Confidence intervals for Dyslexia Index population means

The summary table below presents a comparative overview of summary data for the research groups and subgroups that includes 95% and 99% confidence interval values for the respective population mean Dx values.

confidence intervals for population means

Principal Component Analysis, effect sizes and t-test summary

The summary table below presents an outcome summary of PCA conducted on the two principal metrics: ABC and Dx - refer to the StudyBlog for a more comprehensive commentary but briefly, the table shows the ABC effect sizes between research subgroups ND-400 and DI-600 but particularly between research subgroups DNI and DI-600.

The effect size values are Hedges 'g' results obtained between ABC mean values when the datasets are sifted according to the Dyslexia Index Factor criteria. For example, looking at the row of data for Dx20 5-factor 2: Thinking & Processing, the effect size value of 0.43 shown in the column for ABC24 5-factor: Engagement, indicates the effect size between the ABC means for the factor 'Engagement' when the datasets are sifted according to Dx values established ONLY according to the Dx factor 'Thinking & Processing'. The effect size value of 0.43 in this grid square does NOT indicate the effect size between the Dx factor 'Thinking & Processing' and the ABC factor 'Engagement'. Refer to the StudyBlog for a more detailed explanation of this results summary table.

effect size results from PCA

Dyslexia Dimensions

The summary table below presents a summary of differences in the mean values for each individual dimension of dyslexia which, when taken together, consistute the Dyslexia Index metric devised for this project. These are the means of values returned respondents in the project against each of the dimensions.

The summary presents data for research group ND - that is, students who disclosed no dyslexia; research group DI, students with dyslexia; research subgroup ND-400, this being the subgroup of respondents from research group ND whose Dyslexia Index fell below Dx = 400; research subgroup DI-600, this is the subgroup of respondents from research group DI whose Dyslexia Index was above Dx = 592.5; and finally the research subgroup of real interest, DNI, which is respondents from research group ND whose Dyslexia Index was also above Dx = 592.5 - that is, students presenting dyslexia-like profiles who were not previously identified as dyslexic.

Data shown are the mean values for each dimension together with the respective sample standard deviations. Also shown are the p-values from Student's t-Test for significant differences between independent sample means (equality of variances assumed by Levene's Test, one-tail test at 95% significance level) and finally, the effect size between the means calculated using Hedge's 'g'.

Results of particular interest are in shaded cells, these being where the t-Test p-value indicates at least a significant difference between the sample means, or where Hedge's 'g' is indicating at least 'moderate' effect size.


Differences in mean value dyslexia dimensions


When I was learning to read at school, I often felt I was slower than others in my class My spelling is generally (good) weak* I find it very challenging to manage my time effectively I can explain things to people much more easily verbally than in my writing I think I'm a highly organized learner In my writing I frequently use the wrong word for my intended meaning I generally remember appoint-ments and arrive on time When I'm reading I sometimes read the same line again or miss out a line altogether I have difficulty putting my writing ideas into a sensible order In my writing at school, I often mixed up similar letters When I'm planning my work I use diagrams or mindmaps rather than lists or bullet points I'm hopeless at rememb ering things like telephone numbers I find following directions to get to places quite straght forward I prefer looking at the big picture rather than focusing on the details My friends say I often think in unusual or creative ways to solve problems I find it really challenging to make sense of a list of instructions I get my lefts and rights easily mixed up My tutors often tell me that my essays or assignments are confusing to read I get in a muddle when I'm searching for learning resources or information I get really anxious if I'm asked to read out loud  
* this dimension read '... good' in the research QNR and data was subsequently reverse coded to indicate a level of spelling weakness slow reader spelling manage time explain verbally disorgainzed wrong words for meaning remember appointments miss out lines in reading writing order letters mixedup mindmaps memory following directions big picture innovative problem-solving find lists challenging left-right confusion told essays confusing not systematic anxious reader  
DI means 68.69 67.19 65.43 73.38 44.13 66.40 64.59 83.72 80.00 53.54 51.49 63.09 46.13 64.40 72.46 51.76 64.99 57.10 65.81 77.40  
NDmeans 39.69 35.80 60.02 54.49 46.59 40.93 68.27 53.65 50.85 18.81 37.48 43.99 56.16 53.80 53.89 37.52 35.59 33.47 44.07 44.83  
DI-600 means 78.34 75.45 66.45 84.34 43.32 78.77 68.51 88.26 88.13 67.17 62.51 69.49 42.40 68.47 73.55 58.79 75.28 63.70 74.74 83.38  
ND-400 means 13.05 23.80 52.61 42.34 46.59 21.61 73.43 29.61 27.91 6.52 23.98 27.77 53.39 50.11 44.32 19.18 21.77 14.93 23.98 21.27  
DNI means 73.50 49.17 70.00 79.06 52.72 72.28 63.78 83.83 84.44 45.33 62.06 67.94 65.22 66.50 79.17 65.44 57.78 68.72 78.06 71.78  
t-Test p-values between ND & DI 0.00000 0.00000 0.13728 0.00009 0.29969 0.00000 0.21756 0.00000 0.00000 0.00000 0.00467 0.00026 0.03903 0.00785 0.00001 0.00164 0.00000 0.00000 0.00001 0.00000  
Hedges 'g' between ND & DI 0.7944 1.0359 0.1730 0.6045 -0.0831 0.8688 -0.1235 1.0607 0.9878 1.0700 0.4152 0.5585 -0.2798 0.3853 0.6841 0.4711 0.7978 0.7663 0.7116 0.9654  
t-Test p-values between ND-400 & DI-600 0.00000 0.00000 0.01972 0.00000 0.29743 0.00000 0.20178 0.00000 0.00000 0.00000 0.00000 0.00000 0.07439 0.00069 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000  
Hedges 'g' between ND-400 & DI-600 2.6505 1.9816 0.4385 1.8652 -0.1120 2.7797 -0.1761 2.6920 2.6841 2.1755 1.2349 1.3747 -0.3055 0.6926 1.1330 1.5492 1.6556 1.9271 2.0787 2.3640  
t-Test p-value between DNI & DI-600 0.2836 0.0007 0.3346 0.2012 0.1343 0.1431 0.2801 0.1543 0.1960 0.0138 0.4778 0.4341 0.0173 0.3938 0.2071 0.2078 0.0291 0.2587 0.3056 0.0539  
Hedges 'g' between DNI & DI-600 -0.1594 -0.9299 0.1190 -0.2337 0.3094 -0.2982 -0.1624 -0.2846 -0.2389 -0.6255 -0.0155 -0.0462 0.5986 -0.0750 0.2278 0.2272 -0.5346 0.1805 0.1416 -0.4521