Before recruitment took place it was necessary to ascertain the number of participants required for statistical power.
3.7.1. Determining Participant Numbers for Statistical Power
Statistical power can be defined as the ability of a test to detect an existing effect, i.e. the power to correctly reject a null hypothesis. Power increases when the effect is large and test sensitivity increases with sample size (Laerd Statistics, 2016). In quantitative studies, it is an ethical requirement to pre-determine the sample size required to achieve a statistically significant result without having to inconvenience more participants than necessary. To determine the sample size required for a two-tailed test, a calculation was performed using G*Power (3.1) software (Faul, et al., 2009). The following variables were used:
The significance level (the risk of a Type 1 error, whereby the null hypothesis is rejected when it is true) was set at 5% (p <.05). The power of the study, (the probability that the null hypothesis will be correctly rejected - avoiding a type 2 error) was set at 80%. These are the usual conventions found in research studies of this type.
t tests - Means: Difference between two independent means (two groups)
Analysis: A priori: Compute required sample size
Input:
Tail(s) = Two Effect size d = 0.5 α err prob = 0.05
Power (1-β err prob) = 0.8 Allocation ratio N2/N1 = 1
Output:
Sample size group 1 = 64 Sample size group 2 = 64 Total sample size = 128 Actual power = 0.8014596
Effect size cannot be determined a-priori, however, an estimated or desired effect size can be entered into the calculator. This is commonly set at 0.5 as this provides a
moderate effect size. Choosing a lower value could result in a sample size that produces a statistically significant result having a small or trivial effect size, which would be
undesirable. The calculation shows that a minimum number of 128 participants (64 participants in each group) are required to achieve the statistical significance, and power required.
3.7.2. Sampling Method
Target Population
The target population was intended to be representative of life-long distance learners over the age of 18 in the field of medical or para-medical healthcare education, for example, doctors, nurses, radiographers, physiotherapists, veterinary surgeons, and researchers.
Sampling Method
Participants were invited from a pool of individuals attending study days on the topic of medical imaging. This allowed the data-collection to be performed with the researcher present and assured the participants of full anonymity as the participant ID was only added to the data collection forms post activity. Consecutive sampling was employed. As a non-probability sampling technique, consecutive sampling may not give results that are generalizable to the universal population. However, from all of the non-probability methods, consecutive sampling is considered to give the best generalisability in terms of representing the target population. For this reason consecutive sampling is often used in medical trials (Kendall, 2003). In consecutive sampling, all applicable participants are included in the study until the required number of participants is reached. In other words, every potential participant has a 100% chance of being enrolled in the study. This avoids the types of selection bias that are often associated with non-probability sampling methods such as convenience sampling (Shuster and Powers, 2005). These might include
only sampling participants from a nearby geographical area, or allowing participants to self-select (volunteer) for a study. Consecutive sampling also reduces the opportunity for intentional or unintentional manipulation of the sample, as the researcher is not required to choose participants (Daniel, 2012).
3.7.3. Participant Demographics
The total number of participants was 130, each group containing 65 participants. The pool from which the participants were recruited was wide in terms of age and geographical location.
Age
The participants ranged from 20-64 years (mean = 36.07 years). The mean age of the control group = 36.89 years.
The mean age of the experimental group = 35.26 years.
Geographical Location
Geographically the participants came from the United Kingdom, Ireland, Australasia, the Far East, Middle East, Europe and Scandinavia. All participants were fluent English- speakers.
Gender
The gender ratio overall was 90:40 (69% female: 31% male).
The gender ratio was 46:19 (71% female: 29% male) in the control group. The gender ratio was 44:21 (68% female: 32% male) in the experimental group.
3.7.4. Inclusion Criteria
The inclusion criteria for the study were:
• (All participants) ability to read and write English
• (Experimental group participants) physical ability to manipulate a touch-screen device.
Participants were required to understand English because attempting to undertake the activities without an understanding of English would have introduced an undesirable confounding variable, namely that reading and writing in a second language is likely to have increased task load for reasons other than the learning activity. One participant from the Scandinavian group commented that their English was not of a sufficient standard for the task, and was withdrawn from the study.
3.7.5. Exclusion Criteria
Participants were excluded from the study according to the following criteria:
• Under 18 years of age
• Known to the researcher, or having a professional relationship with the researcher.
The age limit was applied because of ethical considerations and also because the population that the research was intended to generalise for are distance learners who tend to be post-graduates and demographically-speaking 61% of whom are over the age of 30 years (Dabbagh, 2007; Aslanian and Clinefelter, 2012). It was undesirable to use participants who were known to the researcher for reasons outlined in section 3.7.6.
3.7.6. Ethical Approval
An ethics approval form was submitted and approved (see Ethical Approval, p.363). Although the study involves the use of human participants, the data collection methods do not pose any significant risk to the participants. The participants invited to take part in
the study were identified as having no personal connection with the researcher to avoid researcher-bias and to eliminate the possibility of any power relationships between the participant and the researcher. The learners were not students, and therefore had no academic connection with the researcher or the university. This strategy minimised the possibility of coercion or a feeling of obligation to participate and also eliminated potential repercussions on their continued learning, assessment or employability. Recruitment of participants was discontinued as soon as the required sample-size was achieved.
There were no monetary penalties imposed on the participants for taking part as they were already attending the location used for data collection (as part of an educational meeting) and were provided with the materials necessary for the activity. No other incentives were offered to take part in the research.
3.7.7. Informed Consent
Informed consent (see Appendix D :) was sought from all participants before the data collection process and participants were allowed to withdraw consent at any time. The participants were informed about the purpose of the research, the duration of the data collection activity and an outline of the data collection procedure was articulated.
According to the principles of non-maleficence (Dickenson, Huxtable and Parker 2010), it was explained that the procedure did not present any risks, dangers or discomfort to the participants. The possible benefits of the research were also explained, including the potential benefit to software and teaching-material design as well as the small potential benefit that was bestowed by the learning task. The nature of the task was chosen because it related directly to the learners’ own practice.
There was no requirement for deliberate deception in the data collection process (which usually requires a debriefing session), but there was a concluding session that provided the opportunity to thank the participants for taking part. Participants were reminded that there was still an option to withdraw from the study before they submitted their data collection forms.
3.7.8. Confidentiality
All the data collected were stored safely on a secure password-protected non-networked data storage device and were only be accessed by the researcher and project supervisors. No personally identifiable data were collected, and therefore there were no identifiable data that could have been released to a third party. The only exception to this would have been the emergence of data that resulted in the research participant (or others) being put at risk of harm. Due to the nature of the learning task, this did not occur. Participants were assured that in the event of the research findings being published or made public, only aggregate data would be included and that there would be no data published that could identify them as individuals. This fact was also evident from the data collection forms that did not require any identifying information about the participants other than their age and gender.
All data collection, storage and processing complied with the principles of the Data Protection Act 1998 and the EU Directive 95/46 on Data Protection. No personally identifiable data were stored online or on “cloud” based servers.