Quantitative Data Collection Lecture 3 PDF

Quantitative data collection Lecture 3 Dr Owen Hogan Overview  Overview of data collection methods  Primary data collection – surveying, experimental  Secondary data collection - data mining, case studies Primary data collection Surveys and Question...

Quantitative data collection Lecture 3 Dr Owen Hogan Overview  Overview of data collection methods  Primary data collection – surveying, experimental  Secondary data collection - data mining, case studies Primary data collection Surveys and Questionnaires  Surveys and questionnaires are among the most common tools for collecting quantitative data. They consist of a series of questions presented in written form, which can be delivered in person, by mail, online, or over the phone.  They are used to collect data on a wide range of topics, including opinions, behaviours, preferences, and demographic information from a large number of respondents.  Cost-effective, can reach a large audience, standardized responses that are easy to quantify.  Low response rates, potential for response bias, limited depth of information. Structured Interviews  Structured interviews involve asking each participant the same set of predetermined questions in a one-on-one setting. The interviewer follows a strict script to ensure consistency across interviews.  Useful for collecting data when more interaction is needed than a survey allows, but still maintaining the consistency required for quantitative analysis.  Higher response rates compared to surveys, can clarify misunderstandings on the spot, controlled environment.  Time-consuming and more costly than surveys, potential interviewer bias.  Primary data collection Observations  Observational research involves systematically recording behaviours or occurrences from a sample without direct interaction. In quantitative studies, observations are structured and the observer counts or rates specific behaviours or events using a predefined schema.  Ideal for collecting data on behaviour, processes, or events in natural settings.  Provides data in real-world settings, can be less biased by self-reporting issues.  Observer bias, the presence of the observer may alter participants' behaviour (Hawthorne effect). Experiments  Experiments involve manipulating one or more independent variables to determine their effect on one or more dependent variables. This method often involves control groups and random assignments.  Establish causal relationships between variables.  High level of control, can determine causality.  Laboratory settings may not reflect real-world conditions, ethical and practical limitations to what can be manipulated. Primary data collection Longitudinal Studies  These studies involve collecting data from the same subjects repeatedly over a period of time. This could be over weeks, months, years, or even decades.  To track changes over time and analyse trends.  Can identify patterns, trends, and long-term changes, can establish sequences of events.  Time-consuming, expensive, high risk of participant drop-out over time. Electronic Data Collection  Utilizes digital tools and platforms to gather data, such as mobile apps, wearable devices, or online analytics tools. Data collection can be passive (e.g., step counts from a wearable device) or active (e.g., inputs from users).  For collecting real-time data, large datasets, or data that is difficult to capture through traditional methods.  Can collect large volumes of data with minimal effort, real-time data collection, less prone to self-report bias.  Privacy concerns, reliance on technology, potential for data overload.  Secondary data collection Public Records and Archives  Data collected and maintained by governmental agencies, organizations, and institutions, available to the public. These can include census data, birth and death records, and legal documents.  To access comprehensive datasets covering broad populations over extended periods, often used in demographic studies, historical research, and policy analysis.  Usually very reliable and covers extensive periods and large populations.  May not be tailored to specific research needs, and accessing these records can sometimes be challenging due to privacy concerns or bureaucratic hurdles. Official Statistics  Statistics published by government agencies or international organizations, such as employment rates, GDP figures, health statistics, and educational outcomes.  Useful for macro-level research in economics, sociology, public health, and education.  High reliability, generally freely available, covers a wide range of topics.  May not be available for more granular, localized research questions; release frequencies may not match research timelines. Academic, Corporate, and NGO Research  Research findings, datasets, and reports published by academic institutions, corporations, and non-governmental organizations.  To access specialized datasets and analyses that may be costly or time-consuming to replicate, such as longitudinal studies, market research data, or detailed social research.  Often high-quality, subject-specific data; can leverage specialized studies without the associated research costs.  May require purchase or subscription, potential biases depending on the source, the specificity of data might limit broader applicability. Secondary data collection Online Databases and Data Repositories  Online platforms and libraries that store, catalogue, and provide access to a variety of datasets, such as government databases, academic journal databases, and open data repositories like the World Bank Data, Google Public Data Explorer, or ICPSR.  To find and access a wide array of datasets across different fields and subjects, often with tools to assist in data analysis.  Broad access to diverse datasets, often with sophisticated search and analysis tools, increasing the efficiency of secondary data analysis.  Requires digital literacy to navigate and analyse, data quality and relevance can vary widely. Commercial Data Providers  Companies that collect and sell data, often specialized in particular industries or types of data, such as consumer behaviour, media consumption, and financial data.  For access to high-quality, specialized datasets that are not publicly available, particularly in market research and financial analysis.  Comprehensive and detailed datasets tailored to specific industries or research needs.  Can be very expensive, potential issues with data privacy and ethics. Secondary data collection Literature Reviews  Description: Comprehensive reviews of existing academic literature, research reports, and publications that include data relevant to the research question.  Use: To synthesize and analyse data from multiple studies, providing a broader understanding of a research area.  Advantages: Can offer a wide perspective by integrating findings from multiple sources, efficient use of existing data.  Limitations: Time-consuming, relies on the quality and relevance of selected studies, potential publication bias. Media Sources  Description: Information and data available from news outlets, magazines, broadcast media, and digital news platforms.  Use: For contemporary or historical data on public opinions, trends, and events.  Advantages: Easily accessible, can provide insights into public opinion and social trends.  Limitations: Potential biases, accuracy may vary, data may need extensive processing to be usable quantitatively.  Using secondary data can greatly enhance a research project by providing access to high-quality data sets, saving time and resources. However, researchers must critically evaluate the sources of secondary data to ensure their reliability, validity, and relevance to the research question. Validated survey instruments  Validated surveys – have undergone rigorous testing to ensure consistency and stability of measurement. They measure the constructs more accurately and have been validated against established criteria.  Shim, Xiao, Barber, and Lyons (2009) - The Financial Behaviour Scale assesses various dimensions of financial behaviour, including positive and negative financial behaviours.  Non-validated – may lack reliability and validity and lead to inaccurate measurement of the constructs. You need to acknowledge limitations and potential biases and provide enough detailed description of the methodology to allow peers to assess the robustness of the survey  Ball and Shivakumar - The Impact of Stock Option Expensing on Investors' Perceptions and Analysts' Earnings Forecasts": Class activity – Constructing a survey  RQ - How does Taylor Swift's music influence the emotional, cognitive, and behavioural aspects of students' lives, and what is the overall impact of her music on their personal and academic well-being?  Validated survey instrument: The Music Impact Rating Scale, Hays and Minichiello (2005)  Access Google form here: https://docs.google.com/forms/d/1I0HSWwYR1kfPRyNVrEXXb5y2OUEnwwsGv9R- XMqDplo/edit?pli=1