• No se han encontrado resultados

Aplicación de la sonda berberina a un sistema de electroforesis en gel

MOLECULARES PROPUESTAS A SISTEMAS DE CROMATOGRAFIA EN CAPA FINA DE ALTA

5.3. Aplicación de la sonda berberina a un sistema de electroforesis en gel

An important consideration in the development of speaking test tasks is the length of planning time that is provided before the task. In the literature, the length of planning time is a key difference between studies with a task-based language teaching (TBLT) focus and those with a language testing focus. TBLT studies frequently report gains in complexity, accuracy and fluency (CAF) under a ten-minute planning condition (Bui and Huang, 2016, Foster and Skehan, 1996, 1999, Geng and Ferguson, 2013, Kawauchi, 2005, Pang and Skehan, 2014, Skehan and Foster, 1997, 2005, Yuan and Ellis, 2003). Gains in CAF have also been reported under a five-minute pre-task planning condition (Mochizuki and Ortega, 2008, Philp et al., 2006, Sasayama and Izumi, 2012) and under a fifteen-minute planning condition (Sangarun, 2005) with similar results to studies that use ten minutes planning time.

In the language testing literature, the amount of planning time is consistently less than ten minutes (one minute has featured most frequently, Li et al., 2014, Weir et al., 2006, Wigglesworth, 1997, Xi, 2005, 2010) and the effect of the pre-task planning on the test result is inconsistent. Researchers have argued that the inclusion of planning time in a language test has a bearing on test practicality (Elder and Wigglesworth, 2006). Large scale, high stakes testing is often carried out under stringent time constraints. Including a period of ten minutes planning per language test would increase the amount of time required to assess every test taker. The studies with a language testing focus tend to account for this by investigating the impact of short amounts of planning time. However, short amounts of planning time have not been shown to have as substantial an impact on results as planning for ten minutes. This suggests that there may be an optimal amount of planning time that needs to be included in the test to bias for the best performance (Swain, 1985).

Despite variation between research findings, relatively few studies have compared task performance after different amounts of pre-task planning time: Elder and Wigglesworth (2006), and Li et al. (2014). These studies present conflicting accounts of the effect of variation in planning time. Elder and Wigglesworth (2006) found no difference between test scores (rater scores and CAF) after no pre-task planning, under a one-minute planning condition, and under a two-minute planning condition. However, in a questionnaire about their use of the pre-task planning time, test takers expressed a preference for planning. The researchers suggest that as two minutes did not benefit the test takers in terms of test scores, extending planning time on the IELTS exam beyond one minute would be redundant.

Li et al. (2014) used measures of CAF (see Section 2.7.2) to compare the effect of five periods of planning time: 30 seconds, one minute, two minutes, three minutes, and five minutes. Overall, the increases in planning time led to progressively more accurate language (i.e. more error free analysis of speech units (AS-units) and fewer errors per AS-unit). The five-minute planning condition resulted in the most accurate performances overall. However, the largest gains were made between the 30 seconds and one-minute planning conditions, where the number of error free AS-units increased from .48 to .60. Fluency was measured with mean length of run (mean number of syllables supplied between pauses above 0.28 seconds), and speech rate A (syllables per minute) and speech rate B (meaningful syllables per minute). Mean length of run and speech rate A showed incremental increases with every addition of extra planning time up to three minutes. The increases peaked at three minutes and results were lower under the five-minute planning condition. The researchers suggest that this result is evidence of an optimal planning condition for fluency (i.e. three minutes) which, if exceeded causes the effect of pre-task planning to decrease. This conclusion was not confirmed by the speech rate B results, which increased with every addition of extra planning time, i.e. five minutes led to the highest results when only the ‘meaningful’ syllables were calculated in the analysis (Li et al., 2014, p. 46). Complexity results showed that 30 seconds planning led to the highest levels of syntactic complexity and one minute led to the highest level of lexical complexity. This was a surprising result. The researchers do not provide any explanation for the results of the lexical analysis. However, based upon the results of previous research findings (e.g. Crookes, 1989, Yuan and Ellis, 2003), Li et al. suggest that their planning conditions did not provide sufficient time to raise syntactic complexity, and

that ten minutes pre-task planning may be necessary for increases in syntactic complexity to occur.

In sum, the findings reported in Li et al. (2014) indicate that increasing the amount of planning time does not cause systematic increases in all measures of CAF. Perhaps most importantly, the researchers suggest that three minutes is the optimal period of planning for eliciting high levels of fluency. Planning for periods in excess of this amount may cause the planning impact to diminish. This finding does not correspond to the broader research findings reported in the TBLT research, which show that ten minutes planning consistently leads to increases in measures of speech fluency. Nor does it support the results of language testing research that has investigated planning for three minutes (Elder et al., 2002, Elder and Iwashita, 2005, Nitta and Nakatsuhara, 2014). In the literature, the optimal amount of planning time is ten minutes (Ellis, 2009).

The inconsistency in the reported findings of Elder and Wigglesworth (2006) and Li et al. (2014) may perhaps have resulted from differences in the research settings. The former was conducted in Australia (i.e. an English as a second language context; see Section 2.4.1), whereas Li et al. was conducted in China (i.e. an English as a foreign language context). Li et al. involved a homogenous group of participants with similar levels of ability in English that did not have the opportunity to use the language frequently and so develop spoken language proficiency. Proficiency has been identified as a key variable in the effect of pre-task planning (see Section 2.6.2). Additionally, it is not immediately clear that the research participants in Li et al. were aware that they were being tested. The study was conducted in a language laboratory

with participants speaking directly to a computer and there is no indication in the research that the participants were informed that the speaking samples would be used to assess their L2 ability. In contrast, Elder and Wigglesworth investigated the impact of planning on IELTS, a high stakes exam used to determine eligibility to follow English-medium education and for purposes of immigration. The literature indicates that the impact of planning on spoken performance varies substantially under examination and non-examination conditions (see Tables 1 and 2): benefits are more frequently observed when the participant is not being tested. In short, test taker characteristics and the social setting may account for much of the disparity between the studies.

A major gap in the literature is the absence of research both in language testing and TBLT that compares the effect of ten minutes planning with other lengths of planning time. This is problematic; ten minutes has most consistently led to positive results in the TBLT research. The most crucial comparison to make is between test scores under a ten-minute planning condition and under a one-minute planning condition. This is because the studies that investigate the impact of planning for one minute have been conducted with a language testing focus and show inconsistent results. According to this review, another common length of planning time is five minutes. This length of planning time increased CAF in Tavakoli and Skehan (2005), Li et al. (2014), Mochizuki and Ortega (2008), and Sasayama and Izumi (2012).Five minutes may be a sufficient increase over three minutes, which has consistently been shown to have a minimal impact on test scores and CAF measures in the literature (Elder et al., 2002, Elder and Iwashita, 2005, Iwashita et al., 2001, Nitta and Nakatsuhara, 2014). To obtain results after a short period of planning, the research

commonly uses 30 seconds planning time, which is generally deemed sufficient for test takers to familiarise themselves with task demands (Elder et al., 2002, Iwashita et al., 2001, Li et al., 2014).

2.5.2.1 Summary

This section has described research that compares the effects of different lengths of pre-task planning time. To sum up, planning for ten minutes before a language task most consistently leads to high levels of CAF in TBLT. This is in contrast to studies with a language testing focus, which due to practicality constraints, investigate the effect of less planning time (typically one minute), and report inconsistent results. Elder and Wigglesworth (2006) show that test scores are similar after no planning, one-minute, and two-minute planning conditions. However, Li et al. (2014) report differences in CAF after minor changes to planning conditions (e.g. from 30 seconds to one minute). The amount of planning time may be an important variable in the results of the research. The present study investigates the impact of four pre-task planning conditions on test scores: 30 seconds, one minute, five minutes, and ten minutes.