Scientific research methods form the foundation for informed decision-making in Food and Beverage Management. Applying these methods enables professionals to systematically investigate industry challenges, improve service quality, and innovate product offerings in Kenyan hospitality settings. Understanding different types of scientific study designs equips managers and researchers with tools to select appropriate approaches that yield valid and reliable findings. This chapter focuses on the main types of scientific study designs relevant to food and beverage contexts, providing a framework for rigorous research and evidence-based practice.

2.1 Types of scientific study design

A study design is the overall framework that sets out how a research question will be answered: which variables are measured, whether the researcher changes anything, how groups are compared and when data is collected. Choosing the right design is crucial, because it decides what conclusions the study can honestly draw. A descriptive design can say what is happening, a correlational design can say which things move together, and only an experimental design can show that one thing causes another. Example: a manager at a highway restaurant in Mwingi first describes who stops to eat, then checks whether time of day is related to average spend, and finally tests whether a new combo meal raises sales.

Descriptive research design

A descriptive research design systematically describes the characteristics of a population, situation or event as it is, without manipulating any variable or trying to explain causes. It answers questions of who, what, where, when and how many, and provides baseline facts on which managers plan and on which later studies build. Data is usually collected by questionnaires, observation or record review and summarised with frequencies, percentages, means and charts.

Types of descriptive design

• Survey: A structured questionnaire or interview given to a sample to describe opinions, behaviour or characteristics, such as the dining preferences of conference guests.
• Case study: An in-depth description of one unit, such as a single restaurant's approach to reducing food waste.
• Cross-sectional study: Data collected once, at a single point in time, to show the situation at that moment, such as hygiene ratings of eateries in one town in one month.
• Observational study: The researcher watches and records behaviour as it happens, such as how long guests wait to be seated during lunch.
• Developmental or trend study: The same characteristics are described at intervals to show change over time, such as monthly menu item sales over a year.

Applications in food and beverage management

• Profiling guests: Identifying the age, origin, income and spending patterns of restaurant patrons to target marketing.
• Menu analysis: Documenting the range, prices and popularity of dishes to guide menu engineering.
• Service quality assessment: Measuring guest satisfaction levels to benchmark performance against competitors.
• Inventory tracking: Recording the types and quantities of supplies used over a period to improve purchasing and reduce waste.
• Operational mapping: Charting workflows in a commercial kitchen to identify bottlenecks.

Strengths and limitations

Descriptive designs are relatively quick and inexpensive, can cover large samples and give a clear picture of current conditions. However, they cannot explain why something happens or prove cause and effect, and their accuracy depends on honest responses and a representative sample.

Correlational research design

A correlational research design measures two or more variables as they naturally occur and determines whether, and how strongly, they are related. The researcher does not change anything; he or she collects data on both variables from the same subjects and calculates a correlation coefficient. It is used when an experiment would be impossible, unethical or too costly, and when a manager wants to know which factors are worth investigating further.

The correlation coefficient

The strength and direction of a relationship are expressed by the correlation coefficient (r), which ranges from minus 1 to plus 1.

• Positive correlation (r above 0): Both variables rise together, for example the more courteous the waiter is rated, the higher the tip.
• Negative correlation (r below 0): One variable rises as the other falls, for example the longer the waiting time, the lower the satisfaction score.
• Zero or near-zero correlation: No consistent relationship, for example between a guest's shoe size and choice of dessert.
• Strength: Values from about 0.7 to 1.0 (or minus 0.7 to minus 1.0) are strong, 0.4 to 0.69 moderate, and below 0.4 weak.
• Perfect correlation (r = 1 or r = minus 1): All points lie exactly on a straight line; this is rare in real hospitality data.

Example: a coffee house in Rongo records daily rainfall and daily hot beverage sales for three months and finds r = 0.68, a moderate positive relationship.

Uses in food and beverage management

• Service and satisfaction: Relating waiting time, staff courtesy or cleanliness ratings to overall guest satisfaction.
• Pricing and demand: Relating menu prices to the number of portions sold.
• Staffing and output: Relating the number of staff on duty to covers served per hour.
• Training and performance: Relating hours of staff training to upselling results.
• Forecasting: Where a strong relationship exists, one variable can be used to predict the other, for example occupancy to predict breakfast covers.

Correlation is not causation

A correlation shows that two variables move together, not that one causes the other. A third variable may cause both, or the direction may be the reverse of what is assumed. For instance, ice cream sales and soft drink sales are strongly correlated, but neither causes the other: hot weather drives both. Cause and effect can only be established with an experimental design.

Experimental research design

An experimental research design is one in which the researcher deliberately changes (manipulates) one variable, the independent variable, under controlled conditions and measures its effect on another, the dependent variable. Because other factors are held constant and subjects are assigned to groups at random, any difference in the outcome can be attributed to the change. It is the only design that can establish cause and effect.

Characteristics of experimental designs

• Manipulation of variables: The researcher deliberately changes the independent variable, such as the cooking method, portion size or menu layout, to assess its impact on product quality or guest choice.
• Control group: A group that does not receive the treatment serves as a baseline, so that observed effects can be attributed to the treatment rather than to other factors.
• Randomisation: Subjects or test units are assigned to the experimental and control groups at random to reduce selection bias.
• Control of extraneous variables: Other factors that could affect the result, such as day of the week or the waiter on duty, are kept the same for both groups.
• Replication: The experiment is repeated to confirm that the results are consistent.
• Blinding: Subjects, and sometimes researchers, do not know who received the treatment, for example in a taste panel comparing two recipes coded only as A and B.

Types of experimental design

• True experiment: Random assignment to experimental and control groups, with pre-test and post-test measurement.
• Quasi-experiment: Groups are compared without random assignment, for example two branches of a restaurant chain, one of which adopts a new procedure. It is common in hospitality where random assignment of guests is not practical.
• Pre-experiment (one group pre-test, post-test): One group is measured before and after a change, with no control group; it is simple but weak.
• Field experiment: Carried out in a real setting such as a working restaurant.
• Laboratory experiment: Carried out in a controlled setting such as a sensory evaluation room in a training kitchen.

Steps in conducting an experiment

1. State the hypothesis: For example, "Displaying calorie information on the menu reduces orders of deep-fried dishes."
2. Identify the variables: The independent variable (calorie labelling) and the dependent variable (orders of deep-fried dishes).
3. Select and assign subjects: Assign tables, shifts or days to the experimental and control groups at random.
4. Control extraneous variables: Keep prices, dishes and service staff the same for both groups.
5. Apply the treatment: Use the labelled menu for the experimental group only.
6. Measure the outcome: Record orders for both groups over the same period.
7. Analyse and conclude: Compare the groups with a suitable statistical test to accept or reject the hypothesis.

Example: an institutional kitchen at a sugar factory staff canteen in Migori serves one batch of githeri with reduced salt and one with the usual salt on alternate days, and compares the plate waste and staff ratings of the two.

Time dimension of study designs

Any of the designs above can also be classified by when the data is collected, which affects what change over time the study can show.

• Cross-sectional: Data collected once from a sample at one point in time, giving a snapshot, such as the frequency of foodborne illness reports across eateries in one month.
• Longitudinal or cohort: The same group is followed over months or years to see change, such as tracking the career progress of a cohort of hospitality graduates.
• Retrospective (case-control): Starts from an outcome and looks back for causes, such as comparing guests who fell ill after a function with those who did not, to find the food responsible.
• Prospective: Starts now and follows subjects forward to see what outcomes develop.

Comparison of the three designs

Chapter Summary

This chapter explored various types of scientific study design, emphasizing their role in structuring research to ensure reliable and valid results. It examined research methodology, detailing the systematic processes that guide investigators in planning, conducting, and analyzing studies. The discussion on sampling techniques highlighted methods for selecting representative subsets of populations, which are crucial for generalizing findings accurately. Data collection tools were reviewed, focusing on instruments and approaches that facilitate the gathering of quality information necessary for meaningful analysis. Ethical considerations in research were addressed, underscoring the importance of protecting participants' rights and maintaining integrity throughout the research process. Together, these elements form the foundation of rigorous scientific inquiry in hospitality research, enabling practitioners to generate evidence that informs decision-making and improves industry practices.