SPSS Labs - BHAN334 Part 1 PDF

SPSS Labs Variables and Research Data ● Variables are characteristics that assume different values (research data) for different people, objects, or events being studied ● Independent - manipulated by the researcher (presumed cause/input) ● Dependent - outcome (presumed effect/output) ○ Variables ca...

SPSS Labs Variables and Research Data ● Variables are characteristics that assume different values (research data) for different people, objects, or events being studied ● Independent - manipulated by the researcher (presumed cause/input) ● Dependent - outcome (presumed effect/output) ○ Variables can be one of the four measurement scales: nominal, ordinal, interval, ratio ● Measuring information on variables results in research data (data collection) ● Research data help us perform statistics ● Note: Research data can also be referred to as simply data; data collected from a sample represents a data set; other names for data are scores, points, values, observations Types of data ● Numerical (ratio/interval) ○ Quantitative data: BP, weight ● Discrete ○ can be counted, finite ○ A test score is measured in discrete units of 0, 1, 2 ● Continuous ○ can not be counted, infinite ○ the time required for a child to complete a task is continuous and can take any value greater than 0 ● Categorical (nominal) ○ Qualitative or Yes/No data: gender, color ● Ordinal ○ Mixes numerical and categorical data (rating scales Data entry in SPSS ● Variable View ○ Where we define all variables ○ One variable per row (left to right) ● Name ○ unique; must start with a letter; may have up to 8 characters including letters, numbers, or underscore but no spaces (correct: no, id_no; incorrect: id no) ● Type ○ Numeric (numbers only) or string (combination of numbers and letters) ● Width ○ Number of characters SPSS will allow to be entered for each variable (8) ● Decimals ○ Number of decimals per numeric entry (2) ● Label ○ Text to identify in more detail what a variable represents (idno: identification number; or idno: what is your identification number?) ● Values (coding) ○ A way to change qualitative data to quantitative data ○ Assigns a numerical value to a qualitative trait ○ For categorical data ■ Gender: male – 1; female – 2 ■ Dosage: placebo – 1, low dose – 2, high dose – 3 ○ For continuous data: None (no coding needed) ● Missing data ○ None or 999 (for those cells/question items where data was not reported insert 999) ○ Example: “male” -1, “female” – 2, “missing” - 999 ● Columns ○ how wide the column should be for each variable (how much space) (8) ● Align ○ left-justified, right-justified or centered ● Measure ○ Nominal ■ No order, just naming or categorizing (examples: gender, favorite color, religion, marital status) ■ Nominal data is also referred to as “categorical data” ■ A special type of categorical data are “dichotomous” data (2 categories; examples: male and female, pass and fail) ○ Ordinal ■ Allows rank order (examples: educational level – high school degree, undergraduate degree, graduate degree; satisfaction level - very dissatisfied, dissatisfied, somewhat satisfied, satisfied, very satisfied) ○ Scale (interval/ratio) ■ Interval: equal intervals between items; no true zero (examples: temperature, SAT, GRE) ■ Ratio: equal intervals btw items, true zero (examples: distance, time, height, weight, age) ● Role ○ input Create a Codebook ● It can be a copy of the Variable View (in either word or excel format) showing the name, label, values and measurement level for each variable ● Shows how you coded each variable How to get a codebook in SPSS ● In SPSS: File – Display Data File Information – Working File ● (2 tables: Variable Information and Variable Values) Data View ● Where we insert data from every survey ● Each column represents the data collected per one variable ● Each row represents data collected per one respondent/observation Data Distribution ● Refers to the number of times (frequency) each possible observation/outcome occurs in a data set ● There are different ways (numerical and visual) to describe the distribution of data ● Visual ○ Histograms ○ Bar charts ○ Stem and leaf plots ○ Box plots ● Numerical ○ Measures of central tendency ○ (mean, median, mode) ○ Measures of variability ○ (range, variance, standard deviation) Measures of Central Tendency ● Describes a set of data by identifying the central position within that set of data (a.k.a. summary statistics) ● Mean – the sum of a set of numbers divided by the number of respondents/observations. ○ Suppose we have a sample of 7 participants with the following test scores: 2, 2, 4, 1, 2, 3, 3 ■ Mean =(2+2+4+1+2+3+3)/7=2.43 ● Mode – the most frequently occurring value. There can be multiple modes. ○ Mode =2 ● Median – the middle value of a set of numbers (for odd numbered samples); the mean of the middle two points (for even samples). ○ 1,2,2,2,3,4 ■ Median = (2+2)/2=2 Measures of Variability (Spread) ● Indicate the degree to which data points in a distribution are spread out ● Range - The difference between the highest and the lowest data points in a data set ○ Range = 4-1 = 3 ● Variance - The average of the squared differences from the Mean ○ Variance = 5.714/7 = 0.816 ○ X–test scores; M-mean, N-number of participants ● Standard Deviation - The average distance of each point from the Mean ○ SD = square root of the variance = 0.903 ○ Each data point is on average 0.903 points away from the mean SPSS functions for data distribution ● Descriptives ○ Analyze – Descriptive Statistics – Descriptives (move all variables of interest in the Variable(s) box – Options (click Mean, Range, Std Deviation, Maximum and Minimum) – Continue ■ Descriptive Statistics table for each variable ● Frequencies ○ Analyze – Frequencies (move all variables of interest in the Variable(s) box) – Statistics (Mean, Median, Mode, Std Deviation, Range, Maximum, Minimum) – Continue – Charts (Bar chart or Pie chart or Histogram) – Continue ■ Statistics and Frequency tables for each variable Normal Distribution ● Considered the most important theoretical distribution in statistics where the mean, median, and mode are equal ● Forms a symmetrical bell-shaped curve (or it is symmetrical about the mean) ○ Example: 1) look at the shape illustrated by the bar chart and 2) calculate the mean, median, and mode ● The Mean = 70, Median = 70, Mode = 70 ● Since all three are essentially equal, and this is reflected in the bar chart, we can assume that the data are normally distributed ● Also, since the median is approximately equal to the mean, we know that the distribution is symmetrical ● Why do we need to know if data from a study is normally distributed? ● Because only then can we use inferential statistics! Statistical tests ● T-Tests ○ Looks at differences between two groups by comparing the means of the groups ○ Assesses whether the means of two groups are statistically different from each other ● Independent samples t test (between-groups) ○ Compares the means between two unrelated groups on a single dependent variable ○ The independent variable should consist of two categorical, independent groups (males vs. females, Caucasians vs. Latinos) ○ Dependent variable must be approximately normally distributed and on the interval or ratio scale of measurement (pounds lost in 6 months) ● Dependent T test (within-groups) ○ Compares the means between two conditions (before and after) within the same group (OR the same group is measured twice across time - time 1 and time 2 on a single dependent variable) ● Independent variable should consist of two categorical, related groups (same subjects are present in both groups; pounds lost in 3 month and pounds lost in 6 months in women/men) ● Dependent variable must be on the interval or ratio scale of measurement and its distribution must be approximately normal Correlation coefficients ● Statistical tests used to measure and describe the strength and direction of a correlation/relationship between two variables ● A correlation is not a causation ● A correlation is also referred to as an association (between 2 variables) ● Pearson correlation coefficient ® ○ Used to examine the correlation between 2 normally distributed interval/ratio variables ○ Tells us the strength (the closest to 1, the stronger the correlation) and the direction (positive + + / - -; or negative + -) of the correlation ○ If r>0 then the correlation is positive; if r<0 then the correlation is negative ● Spearman’s correlation coefficient (rho) ○ Used when at least one variable is not normally distributed, thus both variables are nominal, or when one variable is ordinal and one variable is interval ANOVA (Analysis of Variance) ● A generalization of the t-test to situations where we compare more than two groups ● Tests the variation of an outcome across multiple groups (>2) by comparing the means of those groups ● One-Way ANOVA ○ Used when we compare the means of more than two independent groups ○ The dependent variable is on the interval/ratio scale of measurement and it is approximately normally distributed for each of the groups ○ The independent variable should consist of three or more categorical, independent groups ○ Sample sizes are approximately equal in each group ○ Data/observations on the dependent variable are independent for each group (a subject can not be more than in one group) Focus groups What is a Focus Group? ● A focus group is a small group of six to ten people led through an open discussion by a skilled moderator. The group needs to be large enough to generate rich discussion but not so large that some participants are left out. ● A homogeneous group of strangers comprise the focus group. Homogeneity levels the playing field and reduces inhibitions among people who will probably never see each other again. ● The focus group moderator nurtures disclosure in an open and spontaneous format. The moderator’s goal is to generate a maximum number of different ideas and opinions from as many different people in the time allotted. ● The ideal amount of time to set aside for a focus group is anywhere from 45 to 90 minutes. Beyond that most groups are not productive and it becomes an imposition on participant time. ● Focus groups are structured around a set of carefully predetermined questions– usually no more than 10 – but the discussion is free-flowing. Ideally, participant comments will stimulate and influence the thinking and sharing of others. Some people even find themselves changing their thoughts and opinions during the group. ● It takes more than one focus group on any one topic to produce valid results –usually three or four. You’ll know you’ve conducted enough groups (with the same set of questions) when you’re not hearing anything new anymore, i.e. you’ve reached a point of saturation. A Focus group is not: ● A debate ● Group therapy ● A conflict resolution session ● A problem-solving session ● An opportunity to collaborate ● A promotional opportunity ● An educational session Types of Questions ● Focus group participants won’t have a chance to see the questions they are being asked. So, to make sure they understand and can fully respond to the questions posed, questions should be: ○ Short and to the point ○ Focused on one dimension each ○ Unambiguously worded ○ Open-ended or sentence completion types ○ Non-threatening or embarrassing ○ Worded in a way that they cannot be answered with a simple “yes” or “no” answer (use “why” and “how” instead) ● There are three types of focus group questions: ○ Engagement questions: introduce participants to and make them comfortable with the topic of discussion ○ Exploration questions: get to the meat of the discussion ○ Exit question: check to see if anything was missed in the discussion An Example Questions for a Focus Group on Dental Flossing ● Engagement questions: ○ What is your favorite toothpaste? ○ What do you notice when you look at other people’s teeth? ● Exploration Questions: ○ Who in particular has influenced your dental habits? ○ What are the pros and cons of flossing your teeth? ○ When you floss, how do follow through? When you don’t, why not? ○ How do you feel when told about possible damage caused by not flossing? ○ How do you feel about yourself when you floss regularly? When you don’t? ● Exit question: ○ Is there anything else you would like to say about why you do or do not floss your teeth on a regular basis? ○ Note: Flossers and non-flossers are in separate groups. Recruitment & Participation ● In an ideal focus group, all the participants are very comfortable with each other but none of them know each other. ● Homogeneity is key to maximizing disclosure among focus group participants. Consider the following in establishing selection criteria for individual groups: ○ Gender ○ Age ○ Power ○ Cliques Conducting the focus group ● Ideally, the focus group is conducted by a team consisting of a moderator and assistant moderator. The moderator facilitates the discussion; the assistant takes notes and runs the tape recorder. Focus Group Moderator Traits ● The ideal focus group moderator has the following traits: ○ Can listen attentively with sensitivity and empathy ○ Is able to listen and think at the same time ○ Believes that all group participants have something to offer no matter what their education, experience, or background ○ Has adequate knowledge of the topic ○ Can keep personal views and ego out of the facilitation ○ Is someone the group can relate to but also give authority to (e.g. a male moderator is most appropriate for a group of all men discussing sexual harassment in the workplace) ○ Can appropriately manage challenging group dynamics Assistant Moderator Role ● The assistant moderator must be able to do the following: ○ Run a tape recorder during the session ○ Take notes in case the recorder fails or the tape is inaudible ○ Note/record body language or other subtle but relevant clues ○ Allow the moderator to do all the talking during the group Before the Focus Group Begins ● It may be important to collect demographic information from participants if age, gender, or other attributes are important for correlation with focus group findings. Design a short half page form that requires no more than two or three minutes to complete. ● Administer it before the focus group begins. Establish Ground Rules ● Outline the behaviors that you expect: ○ Talking one at a time ○ Respectful of others opinions ○ No need to agree / disagree ○ No right or wrong answers ○ Be candid and open ○ Create an environment for discussion ○ Horseshoe seating design ○ Give purpose of Focus Group (Consent) Starting the Focus Group Discussion ● Tips ● Before asking the first focus group question, an icebreaker can be inserted to increase comfort and level the playing field. ○ Example: ■ “If you had a limitless budget, where would you vacation?” ■ Write down on a piece of paper the first 3 words that come to mind when you hear... The focus group moderator has a responsibility to adequately cover all prepared questions within the time allotted. S/he also has a responsibility to get all participants to talk and fully explain their answers. Some helpful probes include: ○ “Can you talk about that more?” ○ “Help me understand what you mean” ○ “Can you give an example?” ● It is good moderator practice to paraphrase and summarize long, complex, or ambiguous comments. It demonstrates active listening and clarifies the comments for everyone in the group. ● Because the moderator holds a position of authority and perceived influence, s/he must remain neutral, and refrain from nodding/raising eyebrows, agreeing/disagreeing, or praising/denigrating any comment made. Dealing with Challenging Participants ● A moderator must tactfully deal with challenging participants. Here are some appropriate strategies: ○ Self-appointed experts: “Thank you. What do other people think?” ○ The dominator: “Let’s have some other comments.” ○ The rambler: Stop eye contact; look at your watch; jump in at their inhale. ○ The shy participant: Make eye contact; call on them; smile at them. ○ The participant who talks very quietly: Ask them to repeat their response more loudly. Closing the Focus Group Session ● When the focus group is complete the moderator thanks all participants and distributes the honorarium (incentive). ● Immediately after all participants leave, the moderator and assistant moderator debrief while the recorder is still running and label all tapes and notes with the date, time (if more than one group per day), and name of the group. Analyzing the Data ● In order for all participant comments to be understandable and useful, they must be boiled down to essential information using a systematic and verifiable process. ● Begin by transcribing all focus group tapes and inserting notes into transcribed material where appropriate. ● Clean up transcripts by stripping off nonessential words. ● Simultaneously assign each participant comment/quote a separate line on the page as well as each new thought or idea therein. Label each line with the participant and group number, e.g. a comment from participant 6 in group 2 would be assigned the number 2.6. Compiling the Data ● Each line is then entered into an Excel database as follows: ○ Use a separate Excel data base spreadsheet for each group. ○ Within each spreadsheet, use one sheet per question. ○ Label three columns on each sheet. ■ One column for coding ■ One column for the participant ID# ■ One column for responses ○ Enter each separate response or idea on a separate line with participant ID attached. The coding column is filled in during the next phase -analysis. Analyzing the Data 1. When all comments have been entered, look for common categories or themes across the entries for each question. The most ideal situation is to ask several people to participate in this process. 2. Once consensus has been achieved regarding the best categories for organizing the data, assign a number or letter to each category. 3. Then assign the number/letter of the category that best fits to each entry on the sheet. 4. Use the Excel ‘Sort’ function to group entries by the categories you have assigned to them. 5. If some entries seem inconsistent for their category, consider re-categorizing or adding another category. It may also be apparent that one or more categories can be collapsed. 6. Arrange categories from those with the largest number of entries to those with the smallest. 7. Repeat for each group.