1/81
Looks like no tags are added yet.
New information collected directly to answer a specific research question.
what is primary data
Information that already exists and was collected for a different purpose.
Do not need access to subjects
Can retrieve information otherwise not accessible
faster access
what is secondary data
You have no control over how the data was collected, so you may not know:
Why: Why the data was collected.
When: When it was collected; it may be outdated.
How: How information was measured or defined.
Who: Who was surveyed and if they represent the target population.
Collector: Who collected the data.
Fit: Whether the data meets your research needs.
what are the drawbacks of secondary data
How consistent a measurement is. If you measure something multiple times, you expect to get the same result.
what is reliability
How accurately a measurement measures what it is supposed to measure.
When you measure
something, you want to
measure the correct thing
what is validity
data relevance
what do reliability and validity of secondary data shape
How well the data matches what you need for your research question.
Researchers make sure existing data is useful and matches your current research question.
what is data relevance
converts units of measure
what is data transformation
Comparing information from different sources to make sure it is accurate.
what is data cross checking
identify consumer behaviors: – consumer patterns relevant to the project at hand
trend analysis: include market tracking i.e. observe and analyze trends in the industry
environmental scanning: early indications of market problems and developments, catching problems early
how can you use secondary data to find facts
forecasting sales: combine secondary data to determine when people are most likely to buy
data mining: use computers to analyze big data sets and identify patterns and relationships
neural network: artificial intelligence; computer mimics processing of human brain in decision making
how can you use secondary data to build models
Estimating how big a potential market is using geographic or demographic data.
what is model building
Information collected from inside an organization, such as customer feedback, complaints, accounting records, and sales figures.
what internal data
Information collected and stored by sources outside an organization, such as census data, university research, trade association surveys, and libraries.
what is external data
public entities: Libraries, online journals, newspapers, government reports, and Census information
commercial sources: Companies that sell data, such as financial information, market shares, consumer opinions, purchasing habits, and advertising/media use.
what sources are external secondary data available from (Places outside an organization where you can find data that already exists.)
Nielsen IQ: Provides existing data about retail sales, product trends, market share, and consumer buying habits.
Nielsen: Provides existing data about TV and streaming audiences, media use, and advertising.
what are examples of secondary information
A snapshot of information collected from people who answer questions.
what is a survey
people who take the survey
-draw a sample of respondents from the population
what are respondents
Surveys can collect people's opinions, likes and dislikes, behaviors, reasons for doing things, future intentions, feelings, willingness to pay, and attitudes
what information can you conduct through surveys
Collect large amounts of data quickly.
Low cost: Usually inexpensive to conduct.
Ask questions about many different topics.
Easy to analyze results using statistics.
what are the pros of surveys
Limited control: You cannot control the environment or how much attention people pay.
Errors and bias: People may give inaccurate or biased answers.
what are the cons of surveys
Errors that happen when the wrong people or too few people are surveyed, so the results do not represent the overall population well.
what are sampling errors
errors from mistakes in the design or execution of survey
what are non sampling errors
Unnatural patterns: Errors can create unusual patterns in the data.
Reliability: Errors make it harder to know if the data is consistent and accurate.
Wrong conclusions: Errors can lead researchers to make incorrect decisions.
why do we care about errors in research
The difference between results from a sample and the true results from surveying the entire population.
You survey 100 college students and find that 60% like Starbucks. However, if you surveyed every college student, only 55% might like Starbucks.
The 5% difference is random sampling error.
what is a random sampling error
When the list used to choose people for a survey is missing people or has incorrect information.
Example: A college survey uses a student list that leaves out transfer students, so they cannot be selected for the survey.
what is a sampling frame error
When people who respond to a survey have different opinions or characteristics than people who do not respond.
Example: A college sends out a survey about campus food, but only students who are really unhappy with the food respond
what is a nonresponse error
when individuals choose for themselves whether to join a study, survey, or experiment, rather than being chosen at random by researchers
Example: Only students who love the campus food choose to complete a survey about it, so the results may not represent all students.
self selection bias
Acquiescence: Always answering positively or negatively in a consistent way.
Extremity: Choosing extreme answers, like “strongly agree” or “strongly disagree.”
Interviewer Bias: Giving answers to please the researcher.
Social Desirability: Giving answers that make yourself look good to others
what are reasons for response biases
administrative errors
what are systematic errors
Data Processing Error: Entering or coding data incorrectly.
Sample Selection Error: The sample is chosen incorrectly and does not represent the population.
Interviewer Error: The respondent’s answers are recorded incorrectly.
Interviewer Cheating: Respondent answers are made up or fabricated by research personnel.
what are examples of different systematic errors
interactive: Allows for spontaneous two-way interaction between researcher and respondent
non interactive: Administered in a way that respondent receives a pre- determined set of questions and cannot interact with researcher.
what are the two types of surveys
personal interviews:
Can ask follow-up questions.
Collects more detailed information.
Can record the interview.
Can show products or samples.
Gets complete answers.
Usually has a high participation rate.
Disadvantages:
Takes more time.
Costs more money.
Requires a trained interviewer.
Interviewer bias can affect answers.
what is the first kind of interactive surveys and the advantages/disadvantages
mall intercepts: Researchers stop people at a shopping mall and ask them questions for a survey.
Advantages:
Can ask follow-up questions.
Gets detailed information.
Can show products or samples.
Can get answers quickly from shoppers.
Disadvantages:
May not represent all consumers.
Some people may refuse to participate.
Takes time and money.
Requires trained interviewers.
Interviewer bias can affect answers.
what is the second kind of interactive survey with advantages/disadvantages
door to door interviews - Surveys conducted at a person's home.
Advantages:
Can reach people who are hard to contact, if they are home.
More personal than telephone calls or emails.
Disadvantages:
Difficult to reach people in apartments or places that do not allow soliciting.
Many people will not answer the door.
The sample may not represent the general population.
what is the third kind of interactive survey with advantages/disadvantages
phone interviews- Surveys conducted over the phone.
Advantages:
Respondents may be more honest because they are not face-to-face with the researcher.
Can reach a fairly representative sample.
Interviews can be completed quickly.
Costs less.
Disadvantages:
Do Not Call registries make it harder to reach people.
Not everyone has a landline.
People using cell phones may be distracted or outside the research area.
Many people do not want to talk on the phone.
what is the fourth kind of interactive survey with advantages/disadvantages
mail surveys: Surveys sent to respondents through the mail.
Advantages:
Can reach a wide geographic area.
Less expensive than other survey methods.
Convenient and easy to complete.
Respondents have confidentiality.
Disadvantages:
Often have low response rates (few people return the survey).
Usually have a low completion rate.
Incentives, like small amounts of money or gifts, can increase response rates.
what is the first kind of non interactive survey with advantages/disadvantages
email and internet surveys: Surveys sent to respondents by email, usually through a website.
Advantages:
Very common survey method.
Can send surveys out quickly.
Low cost.
Provides some confidentiality.
Disadvantages:
Employees at some businesses may not feel comfortable answering honestly.
No control over who actually completes the survey.
Can be difficult to break through spam filters.
what is the second kind of non interactive survey with advantages/disadvantages
Situational Characteristics: Things that affect which survey method you choose.
Budget: How much money you have.
Completion time frame: How quickly you need the results.
Data quality requirements: How accurate and reliable the information needs to be.
Task Characteristics: Things about the survey that affect which method works best.
Task difficulty: How hard the questions or activities are for respondents. Example: You cannot do a taste test through an online survey.
Topic sensitivity: How personal or private the questions are.
Interviewer assistance needed: Whether respondents need help answering questions.
Mixed-Mode Survey: Using more than one survey method, such as email and phone interviews, to collect information.
how do you decide a survey method
mixed mode survey because you are using more than one approach at a time
what kind of survey should you conduct if you have a limited budget
Testing a survey with a small group of people before sending it out to find problems with the questions or design.
Can test it with colleagues, clients, or actual respondents.
Helps make sure the survey collects good-quality data.
what is pre testing
Researchers must treat participants fairly and honestly.
Privacy: Protect participants’ personal information.
Avoid deception: Do not trick or mislead participants.
Informed consent: Tell participants the purpose of the research.
Confidentiality: Keep participants’ answers private.
Honesty: Give truthful information.
Objectivity: Avoid letting personal opinions affect the research.
what are ethical issues when looking at final concerns
A research method where the researcher controls the situation to test cause-and-effect relationships.
Causal relationships means finding out whether one thing causes a change in another.
what is an experiment
Constructs: Abstract concepts that can only be measured indirectly, such as customer satisfaction.
Materialism
Creativity
Trust
Brand Loyalty
Price Consciousness
Variables: Measurable things used to measure a construct.
Examples of variables:
Satisfaction score
Likelihood to buy
Frequency of purchase
what are the two kinds of experiment basics
construct: an idea or concept you want to understand. (is it an idea that you cannot directly see or touch)
variable: something you can measure to understand that idea.(can i measure it with a number, answer or observation)
how do you know if it is construct or variable?
Positive relationship: Both variables move in the same direction.
If one increases, the other increases.
If one decreases, the other decreases.
Example: The more someone likes Rose Tea Latte, the more likely they are to buy one.
Negative Relationship: The variables move in opposite directions.
If one increases, the other decreases.
If one decreases, the other increases.
what are the difference between positive and negative relationships
Predict how two constructs (ideas, things that can not be touched) are related, then measure them using variables to test the relationship.
Example:
Construct 1: Customer satisfaction
Construct 2: Purchase intention
Variable 1: Satisfaction rating (1–10)
Variable 2: Likelihood of buying (1–5)
what are experiment basics
The variable that predicts an outcome and can be controlled or changed by the researcher in an experiment.
Example: A researcher tests whether the amount of advertising affects how much people buy.
Independent variable: Amount of advertising.
Easy way to remember: The independent variable is what the researcher changes to see what happens.
what is an independent variable
The outcome the researcher measures to see how it is affected by the independent variable.
Example:
Independent variable: Amount of advertising.
Dependent variable: Willingness to buy a Rose Tea Latte.
what is a dependent variable
relationships between independent and dependent variables
what is a hypothesis
means something can be proven true or false by observing, measuring, or experimenting in the real world
Example: The more someone likes Rose Tea Latte, the more likely they are to buy one.
what is empirically testable
A statement that says there is no relationship between two variables, or that any relationship is due to chance.
what is a null hypothesis
A statement that predicts a significant relationship exists between two variables and that the relationship is not due to chance.
-difference exists between two variables, but do not describe that difference
what is a alternative hypothesis and non directional
Positive Hypothesis: Predicts that two variables move in the same direction, or that one group will have a higher outcome than another.
- Example: Consumers who see a personalized ad will have a greater intention to buy than consumers who see a generic ad.
Negative Hypothesis: Predicts that two variables move in opposite directions, or that one group will have a lower outcome than another.
- Example: Consumers shown a higher price will have a lower intention to buy than consumers shown a lower price.
what is a positive hypothesis vs negative hypothesis
1. Non-Directional Hypothesis: Social media usage is related to an individual's well-being, but the direction of the relationship is unknown.
2. Positive Hypothesis: The more time individuals spend on social media, the more informed they are about what is happening in the world.
3. Negative Hypothesis: The more time individuals spend on social media, the less time they spend communicating with others in person.
Remember:
Non-directional = there is a relationship, but you don't say which way.
Positive = both variables increase together.
Negative = one increases while the other decreases
for social media usage and individuals well being formulate 3 hypothesis
Research participants exposed to the experiment
what are subjects
Different levels of independent variable included in experiment
what are experimental conditions
Experimental group: The group that receives the treatment or change being tested.
Control group: The group that does not receive the treatment and is used for comparison.
what is the difference between experimental and control groups
variable A is related to Variable B
what is a concomitant variable (casual relationship)
Variable A should come before variable B
what is a temporal sequence (casual relationship)
no other variable should be influence relationship between A and B
what is a nonspurious association (casual relationship)
- Correlation: A relationship between two variables. It shows that they are connected, but it does not prove that one causes the other.
Example: The more you like the Rose Tea Latte, the more likely you are to buy it.
- Causation: One variable directly affects or causes a change in another variable.
Example: You like the Rose Tea Latte so much that it causes you to buy it
what is the difference between correlation and causation
Field experiments: Take place in real-life, natural settings.
More realistic because they happen in everyday environments.
Less control over outside factors that could affect the results.
Can be expensive.
Example: Testing a new advertisement in an actual coffee shop.
- Lab experiments: Take place in an artificial or controlled setting.
More control over outside factors.
More control over how the experiment is done.
Less realistic because it is not a normal, everyday setting.
Example: Having people in a research room look at an advertisement and answer questions.
what are the 2 different kinds of experiments
When a company launches a new product in a few selected markets to see how customers respond before selling it everywhere.
what are test markets
Pros
Estimate future sales: Predict how well the product might sell when it launches everywhere.
Test pricing and product features: See which prices, packaging, or product options customers prefer.
Cons
Cost and time intensive: Takes a lot of time and money.
Competitors may find out: Other companies might discover your new product and create a competing product before your full launch.
what are the pros and cons of test markets
An experiment where a company compares two versions of something to see which performs better.
Experimental conditions: The different versions being tested, such as two different website home page designs.
Outcome: What the company measures, such as how many people click, buy something, or complete a purchase.
what is A/B testing
Internal validity: How sure we are that one variable actually caused a change in another variable.
The cause happens before the effect.
The two variables are related.
No other factors explain the result.
- External validity: How well the results of an experiment apply to other people, places, or situations outside the experiment.
what is the difference between internal and external validity (the accuracy and truthfulness of findings)
-History: An outside event happens during the experiment and affects the results..
- Maturation: People naturally change over time.
-Testing: Taking a test or survey before affects how someone performs next time
- Mortality: Participants drop out of the experiment.
- Regression to the mean: Extremely high or low results tend to be closer to average when measured again.
what are some threats to internal validity
Using data from a sample to make conclusions or predictions about a larger population.
Example: Survey 100 Rowan students to estimate how all Rowan students feel about campus dining.
use a small group to make conclusions about a larger group.
what are inferential (not directly stated) statistics
Organizing, summarizing, and presenting data so it is easier to understand.
Example: Making a chart showing the average age of 100 Rowan students.
what are descriptive statistics
tallying up the number of times different data points occur in a set of data.
what is a frequency distribution (descriptive stats)
Mean: The average. Add all the numbers together and divide by how many numbers there are.
Median: The middle number when the numbers are arranged from smallest to largest.
Mode: The number that appears most often
what are the mean, median and mode (descriptive stats) (central tendency)
Summary measures about variables that are calculated using data from a sample.
Example: Calculating the average amount of coffee students drink by surveying 100 students.
what are sample statistics
Summary measures that describe characteristics of an entire population.
Example: The average amount of coffee consumed by every student at Rowan University.
Census information: Data collected from every member of a population.
Often impossible to get population parameters: It can be too expensive or take too much time to collect data from everyone.
what are population parameters
sampling error: Each data point can be biased (for example, surveying 2 off-campus students who both had exams).
Bias introduces error into statistical measurements.
larger samples: Reduce the impact of random error on descriptive statistics.
2 biased answers out of 10 have less impact than 2 biased answers out of 3.
central limit theorem: As sample size increases, the sampling distribution of the mean becomes more normally distributed.
why is sample size so important? (3)
spread of data
distribution-how data is spread around a central point (usually the mean).
-Many statistical tests assume the data follows a normal distribution.
-Shows how likely a score is to happen randomly.
extreme scores: Very high or very low scores that happen less often than average scores.
what is normal distribution
the distance between the smallest and largest values in a set of data.
Easy way to calculate: Largest value − Smallest value.
what is range (dispersion stats)
Measures how spread out the numbers are from the mean (average). , measured in squared units.
- The larger the variance, the more spread out the data is.
what is variance (dispersion stats)
The center or typical value of a set of data.
There are 3 measures of central tendency:
Mean: The average.
Median: The middle number.
Mode: The number that appears most often.
what is central tendency
Shows how far numbers typically spread from the mean, using the original units. It is the square root of the variance.
-- The larger the standard deviation, the more spread out the data is.
what is standard deviation (dispersion stats)
Let's use the numbers 2, 4, 6.
Find the mean: (2+4+6)div3=4)
Subtract the mean from each number: (2-4=-2, 4-4=0, 6-4=2)
Square each answer: (4, 0, 4)
Add the squared answers: (4+0+4=8)
Divide:
how do you find the variance?
High dispersion: The numbers are spread far apart.
Low dispersion: The numbers are close together.
What do Dispersion statistics (variance, standard deviation, range) tell us about distribution