introduction to experimental research
DESCRIPTION
Introduction to Experimental Research. Lawrence R. Gordon Psychology Research Methods I. From last class…Descriptive Stats. DISPLAYING DATA -- distributions, graphs,…”pictures” Properties of distributions – symmetry, modality, skewness Continuing today… - PowerPoint PPT PresentationTRANSCRIPT
Introduction to Experimental Research
Lawrence R. Gordon
Psychology Research Methods I
From last class…Descriptive Stats DISPLAYING DATA -- distributions, graphs,
…”pictures”Properties of distributions – symmetry, modality,
skewness
Continuing today…
SUMMARIZING DATA – measures of central tendency & variability, …”numbers”
Descriptive Statistics
Name some you know of...
What can they do for us?– Tell us more about our data -- Location– Tell us more about our data -- Spread– Help us communicate these to others!
• Highly summarized• Basis of results and tables in reports of studies
Measures of Central Tendency
Mode– the “typical” score– the peak(s) of the frequency distribution
Median– the “center” score– splits distribution into two halves (by area)
Mean– the (arithmetic) “average” of the scores– the balance point of the distribution
(“seesaw”)
Central Tendency, contd.
Pros and Cons --– Mode– Median– Mean
Relationships to distribution shapes --– Symmetric: Mode = Median = Mean– Skewed: Pos (rgt)- Mo < Md < Mn
Neg (lft)- Mn < Md < Mo
Variability
Range– Hi - Lo score– “Goes with” Mode
Interquartile range (IQR)– 75th-%ile - 25th-%ile scores– “Goes with” Median (50th-%ile score)
Standard Deviation (SD)– Sqrt of average of sqrd deviations around Mn– “Goes with” Mean [Handout for reference]
Real Data
First Hour Exam Scores
FIRST HOUR EXAM: Psyc 109, Fall 02
Number Correct out of 50
47.545.042.540.037.535.032.530.027.525.0
FIRST EXAM GRADE (of 50)
Psyc 109 9/26/02
Freq
uenc
y
70
60
50
40
30
20
10
0
Std. Dev = 4.52
Mean = 38.0
N = 233.00
234
1
37.9829
38.0000
37.00
4.51128
20.35164
24.00
24.00
48.00
35.0000
38.0000
41.0000
Valid
Missing
N
Mean
Median
Mode
Std. Deviation
Variance
Range
Minimum
Maximum
25
50
75
Percentiles
Exam I, 9/26/02
Real Data
Baseball salaries revisited
BASEBALL SALARIES 1994
SALARY94
BASEBALL SALARIES 1994
Freq
uenc
y
300
200
100
0
Std. Dev = 1390922
Mean = 1183416.7
N = 747.00
BASEBALL 1994 SALARIES --- Statistics
SALARY94747
0
1183416.68
500000.00
109000
1390921.93
1.93466E+12
1.317
.089
6191000
109000
6300000
170000.00
500000.00
2000000.00
Valid
Missing
N
Mean
Median
Mode
Std. Deviation
Variance
Skewness
Std. Error of Skewness
Range
Minimum
Maximum
25
50
75
Percentiles
Real Data
109 Questionnaire - Scales (briefly!)
COURSE QUESTIONNAIRE CFC SCALE“Concern for Future Consequences” ScaleDescriptive Statistics
Statistics
NC Scale (Range 18 - 90)187
9
62.6390
63.0000
63.00
11.6460
135.6284
-.337
.178
64.00
26.00
90.00
56.0000
63.0000
71.0000
Valid
Missing
N
Mean
Median
Mode
Std. Deviation
Variance
Skewness
Std. Error of Skewness
Range
Minimum
Maximum
25
50
75
Percentiles
CFC Scale (Range 12 - 60)
60.055.050.045.040.035.030.025.020.015.0
CFC Scale (Range 12 - 60)
Freq
uenc
y
60
50
40
30
20
10
0
Std. Dev = 7.26
Mean = 40.9
N = 174.00
Statistics
CFC Scale (Range 12 - 60)174
22
40.9339
41.0000
40.00
7.2553
52.6387
-.325
.184
43.00
17.00
60.00
37.0000
41.0000
46.0000
Valid
Missing
N
Mean
Median
Mode
Std. Deviation
Variance
Skewness
Std. Error of Skewness
Range
Minimum
Maximum
25
50
75
Percentiles
COURSE QUESTIONNAIRE: NC SCALE“Need for Cognition” Scale
Descriptive Statistics
NC Scale Scores
90.0
85.0
80.0
75.0
70.0
65.0
60.0
55.0
50.0
45.0
40.0
35.0
30.0
25.0
NC Scale (Range 18 - 90)
Freq
uenc
y
40
30
20
10
0
Std. Dev = 11.65
Mean = 62.6
N = 187.00
Statistics
NC Scale (Range 18 - 90)187
9
62.6390
63.0000
63.00
11.6460
135.6284
-.337
.178
64.00
26.00
90.00
56.0000
63.0000
71.0000
Valid
Missing
N
Mean
Median
Mode
Std. Deviation
Variance
Skewness
Std. Error of Skewness
Range
Minimum
Maximum
25
50
75
Percentiles
Real Data
My dissertation data A NOTE TO END ON: not everything
interesting in data is captured by a single statistic! We must get to know data well, from many angles, to find its message!
NOW…on to our current topic, Experimental Research...
NOW, RESUMING…
Intro to Experimental Research, A “pseudo-example”… “Time flies when you’re having fun”!
– DOES IT?
A Simple Experiment: “Time Flies”
EXAMPLE: “Time flies when you’re having fun”
Hypothesis: IF one is “having more fun”, THEN time will seem to pass more quickly
Design: • IV: 100 persons randomly assigned to two groups:
– 1: “Having more fun”
– 2: “Having less fun”
• DV: Estimate of a standard 10 minute interval
Procedure: manipulation of cartoon captions
A Simple Experiment, cont.
EXAMPLE: “Time flies when you’re having fun” (cont..)
Results– “Raw data”
– Organized by “frequency distribution”
– Graphs of data• Group “dotplots”
• Group histograms
Quickie summary of results: “More fun” group gave shorter estimates on average than “Less fun” group.
A Simple Experiment, cont.
RESULTS– How describe?
• List of scores
• A picture
• A single number or two?– LOCATION = “central tendency”
– SPREAD = “variability”
“Having Fun” Data
1 6.1 1 10.9 2 13.7 2 13.8 2 9.2 2 14.3 1 15.3 1 9.1 1 8.5 1 10.4 2 8.1 2 8 1 10.5 2 9.7 2 15.7 2 16.5 2 15.5 2 11.5 2 7.1 1 6.5 2 15 1 11.5 1 8 1 11.2 2 9.3 2 12.2 2 13.6
2 12.5 1 10.4 1 7.1 2 14.4 1 11.4 2 13.4 1 6.3 1 7.5 1 9.5 2 17.7 2 10.9 1 7.5 2 6.9 2 11.2 2 21.8 1 6.4 2 13.9 2 11 1 5.8 2 13.2 2 13 1 10.1 1 5.4 2 13 2 4.9 1 10.2 1 8.6
1 11.9 2 13.2 1 8.4 1 8.8 2 10.3 2 12.4 2 12.1 2 8.7 1 11.5 1 6.1 1 10.2 1 4.1 2 13.6 2 13.2 2 14.2 2 16.1 2 18.2 2 15.5 1 10.2 2 11.2 1 7 1 6.2 1 6.7 2 14.3 2 12.8 1 11.3 1 7.8
2 17.1 1 11.4 2 5.3 1 8.5 1 10 2 10.3 1 7.7 1 0.4 1 9.2 1 11.4 1 8.9 1 3.1 1 12.3 2 10.4 1 5.8 1 5.3 2 15.3 2 9 1 11.8
Estimate of 10 minute interval
COND: 2 'Less fun' (No Captions)10
8
6
4
2
0
Std. Dev = 3.35 Mean = 12.5N = 50.00
Estimate of 10 minute interval
COND: 1 'More fun' (Captions)10
8
6
4
2
0
Std. Dev = 2.72 Mean = 8.6N = 50.00
“HAVING FUN” DESCRIPTIVE STATISTICS: HISTOGRAMS W/ LEGENDS
“HAVING FUN” DESCRIPTIVE STATISTICS: TABLES
'More fun' (Captions)Statisticsb
Estimate of 10 minute interval50
0
8.604
8.700
10.2a
2.722
7.408
-.421
.337
14.9
.4
15.3
6.475
8.700
10.600
Valid
Missing
N
Mean
Median
Mode
Std. Deviation
Variance
Skewness
Std. Error of Skewness
Range
Minimum
Maximum
25
50
75
Percentiles
Multiple modes exist. The smallest value is showna.
Experimental Conditions = 'More fun' (Captions)b.
'Less fun' (No Captions)Statisticsa
Estimate of 10 minute interval50
0
12.484
13.000
13.2
3.353
11.240
-.017
.337
16.9
4.9
21.8
10.300
13.000
14.325
Valid
Missing
N
Mean
Median
Mode
Std. Deviation
Variance
Skewness
Std. Error of Skewness
Range
Minimum
Maximum
25
50
75
Percentiles
Experimental Conditions = 'Less fun' (No Captions)a.
A Simple Experiment, cont.
Results (cont.)– Group 1 = “More fun”
• Mean = 8.60, SD = 2.72, N = 50
– Group 2 = “Less fun”• Mean = 12.48, SD = 3.35, N = 50
– Quickie summary of results: the “More fun” group gave shorter estimates of the 10-minute interval, on average, than the “Less fun” group.
NOTE on the Mean & the SD
Most frequently used -- whenever possible Often called “location” and “scale” -- why? Examples of use with NORMAL curve (the
“bell” curve) ---– Why used frequently?– Various commonly used “standard” scales
THAT’S ALL, FOLKS!