Downloading and First Looks at SPSS

class: center, middle, inverse, title-slide

# Downloading and First Looks at SPSS
## <em>No purchasing</em> needed through Salem State Resources
### Tanner Levenhagen
### Salem State University MSIO Program
### 2022-05-31

---

class: middle center inverse

# First we need to get access to SPSS

---

# Salem State offers a "Virtual Desktop" to access SPSS and other tools

.pull-left[

A very popular statistical tool used is called Statistical Package for Social Sciences (SPSS).

It is a software that allows relatively simple data analysis that requires no coding acumen.

]

.pull-right[

Follow this video made by an alum from Salem State, Shahien, to get access to SPSS through the university :

[This link will take you to the video](https://youtu.be/rXcaluXIDwM)

]

<br>

.center[

### Hopefully now you know how to get into SPSS for the future and we can move on to learning about statistics.

]

---

# There are *Descriptive* and *Inferential* Statistics

.pull-left[

### Descriptive
----

- Demographics about the sample observed

- Central tendencies

- Spread of the data

#### Examples of questions answered with Descriptive Statistics:

What are the demographics of my sample? 
Does this intervention have adverse impact on protected groups? 
Is there potential sampling bias present?

]

.pull-right[

### Inferential
----

- Regression and correlation

- Hypothesis testing

- Suggests, supports, or evidence for. **NOT PROVE**.

#### Examples of questions answered with Inferential Statistics:

Are my findings significantly different than the null hypothesis? 
What is the estimated level of **X** in the population? 
Is **X** a 'good' predictor for **Y**?

]

---

# "Thinking Statistically"

.left-column[

#### What it is:

- Probability
- Evidence/Support
- Informative

----

#### What it is not:

- Absolute
- Proofs
- Infallible
- Unchanging

]

.right-column[

### Description, Prediction, Proportion

- Generally, you can frame just about everything in statistics with those three words.

- *Describe* what your sample was with demographics

- *Predicting* a dependent variables from independent variable with regression

- Find the *proportion* or percentage change to endorse a correct answer

]

---

# Statistics in the Wild

.pull-left[

#### You will see statistics in symbol form

`\(r\)` - Correlation coefficient

`\(\rho\)` - Proportion/probability

`\(\beta\)` - Standardized effect size

`\(b\)` - Unstandardized effect size

`\(\chi^2\)` - Test statistic of null hypothesis

`\(\bar{x}\)` - sample mean

`\(\hat{y}\)` - Predicted DV

`\(n\)` - Number of observations

... and many many more!

]

.pull-right[

### Examples in their context:

**Example #1**

Extraversion was significantly correlated with challenge stressor appraisal ( `\(r=\)` .16, p-value< .01).

**Example #2**

...antagonism had three of four unique (negative) effects on performance, even though mediators were present in the model: role ambiguity, `\(\Delta\chi^2 =\)` 5.29, `\(\beta =\)` .08, p `\(<\)` .05; role conflict, `\(\Delta\chi^2 =\)` 7.00, `\(\beta =\)` .10, p `\(<\)` .05; job satisfaction, `\(\Delta\chi^2 =\)` 0.48, `\(\beta =\)` .02, ns; and organizational commitment, `\(\Delta\chi^2 =\)` 5.18, `\(\beta =\)` .08, p `\(<\)` .05. .small[(Chiaburu & Harrison, 2008)]

]

---

# Statistics in the Wild

.left-column[

#### You will see statistics in table form.

#### The goal is to be able to read what the tables and graphs are *actually* telling you without having to read the disscusion.

]

.right-column[

]

---

# Basics to Statistics:

.panelset[

.panel[.panel-name[Central Tendency]

.pull-left[

Many people assume that when others use the word "average" they always are revering the the mean ( `\(\bar{x}\)` / `\(\mu\)` ).

This is not always the case and could very well be referring to other measures of 'central tendency'.

]

.pull-right[

**Mean** ( `\(\bar{x}\)` / `\(\mu\)` )

This statistic is one number that minimizes the 'distance' of all data. It is, in other words, the best guess when there is no other data to go off of.

*What is the height of a random person in the world?*

]

]

.pull-right[

.pull-left[

**Median** ( `\(\tilde{x}\)` )

This is one number given to the middle observation in data. This is much less affected by outliers than *mean* because it moves the median by one observation while mean can be greatly affected.

*What is the median:*

*3, 4, 5, 8, 9, 10*

]

.pull-right[

**Mode**

This does *not* need to be one number and can be in fact multiple numbers. It is the number with the most observations. If two numbers have the same amount than that is 'bi-modal'.

*What is/are the mode(s)*:

*2, 2, 3, 3, 2, 4, 6, 8, 8, 8, 9*

]

]

.panel[.panel-name[Distribution]

.left-column[

Knowing how the data is around the central tendency is even more important in most analysis. Mess around with the 'SD' in the page to the right and press `Plot` to see the difference (don't change the means).

*Why is it important to know the distribution of data during analysis?*

]

.right-column[

]

]

.panel[.panel-name[Skew]

Knowing the spread of your data is important but it is rare that your data will be *normally distributed* or vertically symmetrical. This is important to be able to identify if your data is 'skewed' either positively or negatively.

.pull-left[

### .center[Negatively Skewed]

]

.pull-right[

### .center[Positively Skewed]

Picture

]

]

.panel[.panel-name[Bi-Modality]

.pull-left[

Sometimes your data will seem to have two values that have "humps" in the histogram/line graphs. This can mean many things.

*What are some reasons you think a graph would be bi-modal? Fill in below (press enter for a new line):*

.can-edit[

]

.pull-right[

]

]

]

---
class: inverse

# Recap

- Can access SPSS through Salem State's resources

- Know the difference between descriptive and inferential statistics

- Know what "thinking statistically" means

- Been exposed to what statistics will look like "in the wild"

- Know the basics of statistics

---
class: inverse middle center

# Next Time

### We will get more specific on more relevant statistics we will run into most often in social sciences.

[Link to next lesson]()