Methods of Collecting Data Clause Samples

Methods of Collecting Data. ▇▇▇▇▇▇ and ▇▇▇▇▇▇ (2005) reviewed 56 empirical studies to evaluate how the PC is researched at work. They found that 70% of the studies used cross-sectional questionnaire surveys, 20% used longitudinal questionnaire surveys, and 10% only used qualitative data collected from interviews. Most of the studies consist of large samples, exceeding 200 employees, and are based on self-report questionnaires. Typically, the studies aim to measure the relationship between the PC and other variables, such as work outcomes, using multiple regression and correlation. The content of the PC is usually measured based on a list of items describing possible promises exchanged between the employee and the organisation, and the participants would rate each item on a scale to express the degree to which the employer made that promise. Using this method to collect data allows the researcher to cover a wide range of components comprising the contract that the employee might miss or forget. However, the list of items may not include aspects of the relationship that are important to the employee and could seem arbitrary. Giving an aggregate score for items based on the average ratings gathered from the sample oversimplifies how important each item is to employees (not all components are equally important) and how employees evaluate the PC as human beings. Individuals evaluate the content and satisfaction of the PC in much more complicated ways than using simple arithmetical calculations and do not always operate logically. It is also noted that such studies treat contract fulfilment and breach as dichotomous, or opposites of the same scale. Not many studies represent contract fulfilment and contract breach as two separate variables, allowing employees to represent a state of being satisfied, despite breach. Moreover, self-report questionnaires are problematic in several ways (Bryman, 2012). The participant might not understand some of the questionnaire items and cannot ask for clarification. Most questionnaires rely on multiple-choice-questions or Likert-type scales, which are easier to distort in comparison to open-ended questions. Participants might not take the questionnaire seriously and provide random answers to get it out of the way. They are also subject to acquiescence (saying yes/agreeing to all questions) and/or fixation on a particular end of a Likert-type scale (e.g. choosing ‘3’ for all items). At the same time, most people do not like to write long responses when aske...