Briner et al.
XFMQ902SF
This whole study is rather long, but if you scroll down to page number
64, you'll see that after the cognitive training, there was no
observed difference in raven matrices performace for the
experimental(sWM training) vs. control group.
Also-
"Although Ravens correlated well with both the sWM tasks (p<0.01 for
Blocks
and p<0.001 for Shapes) and with Mathpiles (p<0.01) and it loaded
heavily on the
sWM/gF factor, Figure 23 illustrates that there was no significant
difference in
improvement between the training groups. There was also no significant
difference
between pretest and post-test scores for any group (see Figure 22).
The lack of
improvement on Ravens and its reliance on other cognitive functions,
such as Blocks
and Shapes, suggests that there is another cognitive function utilized
when
completing the task. According to the literature, Ravens is a measure
of fluid
intelligence (gF) (Chein & Morrison, 2010; Westerberg & Klingberg,
2007). This
leads to a transfer effect on gF. Participants received one point for
every correctly
solved matrix."
Gwern Branwen
Reading up on it, span-board looks fairly different from dual or even
single n-back; but the task was adaptive. Hm. Interesting but not
actually n-back; it'll go in a different FAQ section if I ever make a
separate section for 'other' gains.
The full conclusion:
> Our study found transfer effects in all groups to be limited. Both adaptive working memory training and non-adaptive arithmetic training led to improvement on trained tasks, and near transfer to other measures of the trained constructs. The Working Memory group improved on five measures of working memory: two of which were the training tasks, and three of which were untrained tasks. The Math and Reading Comprehension group improved on four measures of math ability: one of which was the training task, and three of which were untrained. This provides further evidence for near transfer of working memory training found by previous studies (Chein et al., 2010). The Working Memory group improved more on the Block Span and Shape Builder tasks than either the Math and Reading Comprehension group or the Low Contact group. This was expected, as the Math and Reading Comprehension group did not train on these tasks and the Low Contact group did not receive as much training as the Working Memory group. Interestingly, both the Working Memory group and the Math and Reading Comprehension group demonstrated more improvement on the Verbal Learning task than the Low Contact group, as the Low Contact group actually decreased performance on all measures of Verbal Learning. The training tasks, as well as the factors they loaded on, correlated well with the Verbal Learning recall variables and the Verbal Learning task factor. This, in combination with the results, led to the conclusion that training on either visuo-spatial working memory tasks or quantitative reasoning tasks can lead to improvement on the recall portion of the Verbal Learning task.
A nice note on Shipstead:
> We utilized a training task similar to the Spanboard task (Olesson, 2004; Klingberg, 2005) in order to target working memory, attentional control and inhibition response; however, our version of the task was made adaptive in order to maximize cognitive engagement and keep participants performing at their cognitive capacity. A battery of pre- and post-tests was selected to assess both baseline performance of participants and gauge changes in performance after training. All confounds listed by Shipstead were considered when designing our methodology, and were controlled for as best as possible.
Other quotes of interest:
> Research has found cognitive deficits in individuals with schizophrenia, suggesting that successful cognitive training may be a valid way of remediating these deficits and allowing schizophrenics to lead more normal lives (Barch, 2005; Lesh, Niendam, Minzenberg, & Carter, 2011; Nestor, Niznikiewicz, & McCarley, 2010; Ranganath, Minzenberg, & Ragland, 2008). Training may also be used to maintain cognitive abilities into old age and ward off dementia (Greenwood, 2007).
> Faille (2006), in a study investigating the effects of cognitive training on the elderly, found that participants with possible mild cognitive impairment outperformed those who were cognitively intact after training. She also found that individuals with no impairment or mild impairment respond similarly to training, and that men and women respond similarly as well.
> Despite the existence of age-related brain atrophy in adults due to synapse loss, dendritic regression, and white matter degeneration, studies suggest that neuroplasticity can facilitate physical recovery in elderly adults (Greenwood 2007). Greenwood (2007) demonstrated that despite the existence of brain shrinkage and atrophy in participants, neuroimaging showed that strategy-induced plasticity training caused increased activity in regions adjacent to the atrophied areas of the prefrontal cortex.
> Chein et al. (2010) found that verbal and visuo-spatial working memory training transferred to measures of inhibition and reading comprehension, but did not lead to measures of abstract reasoning believed to heavily rely on working memory.
--
gwern
http://www.gwern.net