explain how to read this regression model table

Strong payday effects on behavior Payday 1 day before payday 2 days before payday 3 days before payday 4 days before payday 5 days before payday Piece rate increase Lag dependent variable controls Observations R2 Dependent variable mean Production (1) 215 (70)*+* No 8,423 4961 5,337 Production (2) 140 (63)** Yes 8,423 .5889 5,337 DEPENDENT VARIABLE Production (3) 428 (94) %+ 539 (95)** 417 (118)*#= 374 (112)%*= 332 (123)%*=* 176 (119) Yes 8,423 .5909 5,337 Earnings (4) 14.09 (2.99)=+* 17.19 (3.02)*+* 13.54 (3.60) === 11.82 ( 3. 57)*:* 10.15 ( 3.91)%s= 591 (8.79) Yes 8,423 57 172 Attendance (3) 077 (-013)%* 053 (.018)%** 037 (.016)** ((017)ww 023 (017) No 8,423 A1 .88 Production (6) 1,071 (239)*** Yes 550 .76 9,361 NoTte.~The sample in cols. 1-5 is the experiment sample. The dependent variables equal zero if a worker was absent. Payday is a binary indicator for whether that day was the worker's assigned payday. In cols. 3-5, \"X days before payday\" are binary indicators for whether the current day is X calendar days away from the worker's assigned payday; the omitted category in these columns is 6 or more days away from the payday. In col. 6, the sample is observations after the end of the experiment in which workers' wages were randomized. Piece rate increase is a binary indicator that equals one if the worker's piece rate was Rs. 0.04 per accurate field that day and equals zero if the worker's piece rate was Rs. 0.03 per accurate field. All regressions include fixed effects for each date in the sample, each worker in the sample, and each computer seating assignment. Regressions 2, 3, and 6 also include controls for lagged production (production on the previous workday and 2 workdays ago); similarly, regression 4 includes controls for lagged earnings. Robust standard errors are reported in parentheses