Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

9/1/16

An Introduction to
Statistics Psychological
"Statistics  is  a  b it  like  s ticking  your  finger  into  a  
revolving  fan  b lade:  s ometimes  it’s   very  p ainful,  b ut  
Statistics it  d oes  give  you  the  p ower  to  answer  interesting  
questions."  
Andy   Field

Psych  110  – Day  2


1 2

Objective   Outline
• To  define  statistics   and  describe   its  nature.   What  is  statistics?

Basic  v ocabulary  in  statistics


• To  explain   some  basic  concepts   in  statistics.
Data  structures,  research  methods,  and  statistics
• To  appreciate   the  importance   of  statistics   in  
Types  of  v ariables  and  scales  of  measurement
the  study  of  human   behavior.
Statistics  and  Psychology

3 4

1
9/1/16

What  is  statistics?


• STATISTICS  – a  set  of  mathematical   procedures  
for  organizing,   summarizing,   and  interpreting
information.
• Role  of  statistics:
– To  make   sense  of  research  d ata  and  b e  able  to  
communicate  results  to  others.
– To  answer   the  general   research  q uestion(s)  b y   BASIC  VOCABULARY  IN  STATISTICS
determining  what   conclusions  are  justified  b ased  
on  the  results  that  were  obtained.
5 6

Population  vs.  sample The   relationship  b etween  a  p opulation  and  a  s ample

• POPULATION   – the   set   of  all   the   individuals   of   POPULATION


interest   in  a   particular   study. All  o f  t he  individuals  o f  interest

– Can   vary  in  size


– May  not  necessarily   consist   of   people
The   results  f rom  
The   sample   is  
the  sample   are  
selected   from  t he  
• SAMPLE   – a   set  of   individuals   selected   from   a   generalized   to  t he  
population.
population.
population,   usually   intended   to   represent   the  
population   in  a   research   study.
– The  goal  in   examining   samples   is  always  t o  generalize  
the  results  back   to  t he  entire  population. SAMPLE
The  individuals  selected  t o  p articipate  in  
the  r esearch  study.
7 8

2
9/1/16

Variables   and   data Parameter  vs.  statistic


• VARIABLE – a  characteristic  or  condition  that   • PARAMETER – a  v alue  (usually  numerical)   that  
changes  or  has  different  v alues  for  different   describes   a  population;   usually  derived   from  
individuals.
measurements   of  the  individuals   in  the  
– May  be  individual   or  environmental   characteristics
population.
• DATA (pl.)  – measurements  or  observations.
• DATA SET – collection  of  measurements  or   • STATISTIC -­‐ a  v alue  (usually  numerical)   that  
observations. describes   a  sample;  usually  derived   from  
• DATUM (sing.)  – a  single  measurement  or   measurements   of  the  individuals   in  the  
observation;  commonly  called  a  score  or  raw   sample.
score
9 10

Descriptive  vs.  inferential  statistics


• DESCRIPTIVE STATISTICS – statistical   procedures  
used   to  s ummarize,   organize,   and  s implify   data.

• INFERENTIAL STATISTICS -­‐ statistical   procedures  


used   to  s tudy  s amples   and   then   make  
generalizations   about   the   populations   from   which  
they   were   selected.
DATA  STRUCTURES,   RESEARCH  
• SAMPLING ERROR – the   discrepancy,   or   amount  
of  error,   that   exists  b etween   a   sample   statistic   METHODS,  AND  STATISTICS
and   the   corresponding   population   parameter.
11 12

3
9/1/16

• I.  M easuring  2  related  variables  for  each  


individual   The  experimental  method
– CORRELATIONAL METHOD – 2  different   variables  are  
observed   to  determine   whether   there  is  a  relationship   • EXPERIMENTAL METHOD (cont’d)
between   them. – General  categories   of  variables  to  consider:
• 1.  Participant  v ariables  – age,  gender,  intelligence,  etc.
• II.  Comparing  2  (or  more)  groups  of  scores • 2.  Environment  variables  – lighting,  time  of  day,  
weather,  etc.
– EXPERIMENTAL METHOD – 1  v ariable  is  manipulated  
while  another    v ariable  is  observed   and  measured.
– Techniques  in  controlling  variables:
• To  establish  a  cause-­‐and-­‐effect  relationship  between  the  2   • 1.  random  assignment
variables,  an  experiment  attempts  to  control  all  other   • 2.  matching
variables  to  prevent  them  from  influencing  the  results.
• 3.  holding  v ariables  constant
• 2  distinct  characteristics:  manipulation and  control
13 14

The  experimental  method Nonexperimental   methods


• EXPERIMENTAL METHOD (cont’d)
– Types  of   variables  in  an  experiment: • II.  Comparing   2  (or   more)  groups   of  scores
• 1.  Independent variable (IV)  – the  one  manipulated  by  the   – NONEXPERIMENTAL METHODS
researcher. • Nonequivalent  studies  – uses  preexisting  groups;  
• 2.  Dependent variable (DV)  – the  one  observed  to  assess  the   researcher  cannot  control    group  assignment  (e.g.,  sex,  
effect   of  the  treatment. age,  etc.)

– Types  of   experimental   conditions: • Pre-­‐post  studies  – uses  time  (before/after)  t o  define  
• 1.  Control condition/group – individuals  do  not  receive  the   groups
experimental  treatment   or  they  receive   neutral  (placebo)  
treatment;   provides  baseline  for  comparison  w/  
experimental  group. • Quasi-­‐independent  v ariable  – the  IV  that  is  used  t o  
• 2.  Experimental condition/group – individuals  receive  the   create  different  groups  of  scores  in  nonexperimental  
experimental  treatment. studies.
15 16

4
9/1/16

Constructs   &  ODs


• CONSTRUCTS – internal   attributes   or  
characteristics   that  cannot   be  directly  
observed   but  are  useful   for  describing   and  
explaining   behavior.

• OPERATIONAL DEFINITION – describes   a  set  of  


TYPES  OF  VARIABLES  AND  SCALES   operations   for  measuring   a  construct   and  
defines   the  construct   in  terms   of  the  resulting  
OF  MEASUREMENT measurements.

17 18

Types   of  variables Scales/levels   of  measurement


• Types  of  variables  (based  on  their   values): • Scales  (levels)  of  measurement  –set  of  categories  used  for  
assigning  individuals  or  events
– 1.  D ISCRETE VARIABLE – consists  of  s eparate,  
indivisible  categories.  No  values  can  exist  b etween   – 1.  NOMINAL SCALE – set  of  categories  that  have  different  
names.
two  n eighboring  categories. • Labeling   &  c ategorization   only;  no   quantitative   distinctions   between  
observations.
– 2.  ORDINAL SCALE – set  of  categories  organized  in  an  o rdered  
– 2.  CONTINUOUS VARIABLES – divisible  into  an   sequence.
infinite  n umber    of  fractional  p arts. • Measurements   are  ranked  in   terms   of  size/magnitude
– 3.  INTERVAL SCALE – ordered  categories  that  are   all  intervals  of  
• Obtaining  identical  measurements  for  2  different   exactly  the  same  size.
individuals  is  rare. • Equal   differences   between   numbers   =  equal   magnitude
• Each  measurement  category  is  actually  an  interval  t hat   • Zero  point   arbitrary
must  be  defined  by  boundaries  (real  limits) – 4.  RATIO SCALE – an  interval  scale  with  an  absolute  zero  point.
19 20

5
9/1/16

Scales  of  measurement  compared Statistics  and  Psychology


Scale   Level
Scale   of  
Measurement
Scale   Qualities Example(s) • Why  is  Statistics   important   in  Psychology?

Magnitude Age,   Height,  


4 Ratio Equal   Intervals Weight,   • How  to  get  the  most  out   of  Psych  110
Absolute   Zero Percentage
– Read  in  advance
Magnitude
– Ask  q uestions  &  s hare  ideas
3 Interval Temperature
Equal   Intervals – Consistent  s tudy  h abit
Likert   Scale,   – Ask  for  h elp,  esp.  from  one’s  groupmates
2 Ordinal Magnitude Anything   rank  
ordered – Try   not  to  miss  any  meeting
Names,   Lists   of
1 Nominal None
words
– Practice,  p ractice,  p ractice!
21 22

Looking at the larger lesson


Statistics as a more critical way of looking at things

23

You might also like